Giters

AnshMittal1811 / 100_Days_for_ComputerVision_Papers

This is a repo where I keep a track of the Computer Vision papers that I am reading

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

computervisionmachinelearningmachinelearning-pythonpaperresearch

365 Days for Computer Vision Research

Abstract— This work presents a deep learning approach for vehicle detection in satellite video. Vehicle detection is perhaps impossible in single EO satellite images due to the tininess of vehicles (4-10 pixel) and their similarity to the background. Instead, we consider satellite video which overcomes the lack of spatial information by temporal consistency of vehicle movement. A new spatiotemporal model of a compact 3 × 3 convolutional, neural network is proposed which neglects pooling layers and uses leaky ReLUs. Then we use a reformulation of the output heatmap including Non-Maximum-Suppression (NMS) for the final segmentation. Empirical results on two new annotated satellite videos reconfirm the applicability of this approach for vehicle detection. They more importantly indicate that pretraining on WAMI data and then fine-tuning on few annotated video frames for a new video is sufficient. In our experiment only five annotated images yield a 𝐹1 score of 0.81 on a new video showing more complex traffic patterns than the Las Vegas video. Our best result on Las Vegas is a 𝐹1 score of 0.87 which makes the proposed approach a leading method for this benchmark.

Abstract— Salient object detection (SOD) extracts meaningful contents from an input image. RGB-based SOD methods lack the complementary depth clues; hence, providing limited performance for complex scenarios. Similarly, RGB-D models process RGB and depth inputs, but the depth data availability during testing may hinder the model’s practical applicability. This paper exploits only RGB images, estimates depth from RGB, and leverages the intermediate depth features. We employ a pyramidal attention structure to extract multi-level convolutional-transformer features to process initial stage representations and further enhance the subsequent ones. At each stage, the backbone transformer model produces global receptive fields and computing in parallel to attain fine-grained global predictions refined by our residual convolutional attention decoder for optimal saliency prediction. We report significantly improved performance against 21 and 40 state-ofthe-art SOD methods on eight RGB and RGB-D datasets, respectively. Consequently, we present a new SOD perspective of generating RGB-D SOD without acquiring depth data during training and testing and assist RGB methods with depth clues for improved performance. The code and trained models are available at https://github.com/tanveer-hussain/EfficientSOD2

Abstract— Most motion retargeting methods in human action video synthesis decompose the input video to motion (dynamic information) and shape (static information). However, we observe if the dynamic information is directly transferred to another subject, it will result in unnatural synthesised motion. This phenomenon is mainly caused by neglecting subject-dependent information in motion. To solve the problem, we propose a novel motion retargeting method which can combine both subject-independent (common motion content) information from a source video and subject-dependent (individualized identity motion) information from a target video. So it can synthesize videos with a much natural appearance along with identity-preserved motion (Figure 1). In the proposed method two encoders are employed to extract identity and motion content representations respectively. We employ the adaptive instance normalization (AdaIN) layer in the generator and the instance normalization (IN) layer in the motion content encoder to synthesize the new motion. Besides, we also collected a dataset, named𝐶ℎ𝑢𝑎𝑛𝑔101, with 101 subjects in total. Each subject performs identical dancing movement, and so it is convenient for feature disentanglement among motion and identity of each subject. Furthermore, an efficient quantitative metric for identify information is designed by gait recognition. The experiments show the proposed method cansynthesize videos more naturally when the subject’s identity is preserved.

Abstract— Most existing unsupervised person re-identification (ReID) methods use clustering to generate pseudo labels for model training. Unfortunately, clustering sometimes mixes different true identities together or splits the same identity into two or more sub clusters. Training on these noisy clusters substantially hampers the Re-ID accuracy. Due to the limited samples in each identity, we suppose there may lack some underlying information to well reveal the accurate clusters. To discover these information, we propose an Implicit Sample Extension (ISE) method to generate what we call support samples around the cluster boundaries. Specifically, we generate support samples from actual samples and their neighbouring clusters in the embedding space through a progressive linear interpolation (PLI) strategy. PLI controls the generation with two critical factors, i.e., 1) the direction from the actual sample towards its K-nearest clusters and 2) the degree for mixing up the context information from the K-nearest clusters. Meanwhile, given the support samples, ISE further uses a label-preserving loss to pull them towards their corresponding actual samples, so as to compact each cluster. Consequently, ISE reduces the “sub and mixed” clustering errors, thus improving the Re-ID performance. Extensive experiments demonstrate that the proposed method is effective and achieves state-of-the-art performance for unsupervised person Re-ID. Code is available at: https://github.com/PaddlePaddle/PaddleClas.

Abstract— Recent development of neural implicit function has shown tremendous success on high-quality 3D shape reconstruction. However, most works divide the space into inside and outside of the shape, which limits their representing power to single-layer and watertight shapes. This limitation leads to tedious data processing (converting nonwatertight raw data to watertight) as well as the incapability of representing general object shapes in the real world. In this work, we propose a novel method to represent general shapes including non-watertight shapes and shapes with multi-layer surfaces. We introduce General Implicit Function for 3D Shape (GIFS), which models the relationships between every two points instead of the relationships between points and surfaces. Instead of dividing 3D space into predefined inside-outside regions, GIFS encodes whether two points are separated by any surface. Experiments on ShapeNet show that GIFS outperforms previous state-of-the-art methods in terms of reconstruction quality, rendering efficiency, and visual fidelity. Project page is available at https://jianglongye.com/gifs.

Abstract— Following tremendous success in natural language processing, transformers have recently shown much promise for computer vision. The self-attention operation underlying transformers yields global interactions between all tokens, i.e. words or image patches, and enables flexible modelling of image data beyond the local interactions of convolutions. This flexibility, however, comes with a quadratic complexity in time and memory, hindering application to long sequences and highresolution images. We propose a “transposed” version of self-attention that operates across feature channels rather than tokens, where the interactions are based on the cross-covariance matrix between keys and queries. The resulting cross-covariance attention (XCA) has linear complexity in the number of tokens, and allows efficient processing of high-resolution images. Our cross-covariance image transformer (XCiT) – built upon XCA – combines the accuracy of conventional transformers with the scalability of convolutional architectures. We validate the effectiveness and generality of XCiT by reporting excellent results on multiple vision benchmarks, including (self-supervised) image classification on ImageNet-1k, object detection and instance segmentation on COCO, and semantic segmentation on ADE20k

Abstract— —Topic models (e.g., pLSA, LDA, SLDA) have been widely used for segmenting imagery. These models are confined to crisp segmentation. Yet, there are many images in which some regions cannot be assigned a crisp label (e.g., transition regions between a foggy sky and the ground or between sand and water at a beach). In these cases, a visual word is best represented with partial memberships across multiple topics. To address this, we present a partial membership latent Dirichlet allocation (PM-LDA) model and associated parameter estimation algorithms. Experimental results on two natural image datasets and one SONAR image dataset show that PM-LDA can produce both crisp and soft semantic image segmentations; a capability existing methods do not have.

Abstract— We present PHORHUM, a novel, end-to-end trainable, deep neural network methodology for photorealistic 3D human reconstruction given just a monocular RGB image. Our pixel-aligned method estimates detailed 3D geometry and, for the first time, the unshaded surface color together with the scene illumination. Observing that 3D supervision alone is not sufficient for high fidelity color reconstruction, we introduce patch-based rendering losses that enable reliable color reconstruction on visible parts of the human, and detailed and plausible color estimation for the non-visible parts. Moreover, our method specifically addresses methodological and practical limitations of prior work in terms of representing geometry, albedo, and illumination effects, in an end-to-end model where factors can be effectively disentangled. In extensive experiments, we demonstrate the versatility and robustness of our approach. Our state-of-the-art results validate the method qualitatively and for different metrics, for both geometric and color reconstruction.

Abstract— Image-based virtual try-on strives to transfer the appearance of a clothing item onto the image of a target person. Prior work focuses mainly on upper-body clothes (e.g. t-shirts, shirts, and tops) and neglects full-body or lower-body items. This shortcoming arises from amain factor: current publicly available datasets for image-based virtual try-on do not account for this variety, thus limiting progress in the field. To address this deficiency, we introduce Dress Code, which contains images of multi-category clothes. Dress Code is more than 3× larger than publicly available datasets for image-based virtual try-on and features high-resolution paired images (1024 × 768) with front-view, full-body reference models. To generate HD try-on images with high visual quality and rich in details, we propose to learn fine-grained discriminating features. Specifically, we leverage a semantic-aware discriminator that makes predictions at pixel-level instead of image- or patch-level. Extensive experimental evaluation demonstrates that the proposed approach surpasses the baselines and state-of-the-art competitors in terms of visual quality and quantitative results. The Dress Code dataset is publicly available at https://github.com/aimagelab/dress-code.

Abstract— Recent salient object detection (SOD) methods basedon deep neural network have achieved remarkable performance. However, most of existing SOD models designed for low-resolution input perform poorly on high-resolution images due to the contradiction between the sampling depth and the receptive field size. Aiming at resolving this contradiction, we propose a novel one-stage framework called Pyramid Grafting Network (PGNet), using transformer and CNN backbone to extract features from different resolution images independently and then graft the features from transformer branch to CNN branch. An attention-based Cross-Model Grafting Module (CMGM) is proposed to enable CNN branch to combine broken detailed information more holistically, guided by different source feature during decoding process. Moreover, we design an Attention Guided Loss (AGL) to explicitly supervise the attention matrix generated by CMGM to help the network better interact with the attention from different models. We contribute a new Ultra-High-Resolution Saliency Detection dataset UHRSD, containing 5,920 images at 4K-8K resolutions. To our knowledge, it is the largest dataset in both quantity and resolution for high-resolution SOD task, which can be used for training and testing in future research. Sufficient experiments on UHRSD and widely-used SOD datasets demonstrate that our method achieves superior performance compared to the state-of-the-art methods.

Abstract— It has been shown by many researchers that transformers perform as well as convolutional neural networks in many computer vision tasks. Meanwhile, the large computational costs of its attention module hinder further studies and applications on edge devices. Some pruning methods have been developed to construct efficient vision transformers, but most of them have considered image classification tasks only. Inspired by these results, we propose SiDT, a method for pruning vision transformer backbones on more complicated vision tasks like object detection, based on the search of transformer dimensions. Experiments on CIFAR-100 and COCO datasets show that the backbones with 20% or 40% dimensions/parameters pruned can have similar or even better performance than the unpruned models. Moreover, we have also provided the complexity analysis and comparisons with the previous pruning methods.

Abstract— We address the problem of inferring the anatomic skeleton of a person, in an arbitrary pose, from the 3D surface of the body; i.e. we predict the inside (bones) from the outside (skin). This has many applications in medicine and biomechanics. Existing state-of-the-art biomechanical skeletons are detailed but do not easily generalize to new subjects. Additionally, computer vision and graphics methods that predict skeletons are typically heuristic, not learned from data, do not leverage the full 3D body surface, and are not validated against ground truth. To our knowledge, our system, called OSSO (Obtaining Skeletal Shape from Outside), is the first to learn the mapping from the 3D body surface to the internal skeleton from real data. We do so using 1000 male and 1000 female dual-energy X-ray absorptiometry (DXA) scans. To these, we fit a parametric 3D body shape model (STAR) to capture the body surface and a novel part-based 3D skeleton model to capture the bones. This provides inside/outside training pairs. We model the statistical variation of full skeletons using PCA in a pose-normalized space. We then train a regressor from body shape parameters to skeleton shape parameters and refine the skeleton to satisfy constraints on physical plausibility. Given an arbitrary 3D body shape and pose, OSSO predicts a realistic skeleton inside. In contrast to previous work, we evaluate the accuracy of the skeleton shape quantitatively on held-out DXA scans, outperforming the state-of-the-art. We also show 3D skeleton prediction from varied and challenging 3D bodies. The code to infer a skeleton from a body shape is available for research at this https URL, and the dataset of paired outer surface (skin) and skeleton (bone) meshes is available as a Biobank Returned Dataset. This research has been conducted using the UK Biobank Resource.

Abstract— Effectively utilizing the vast amounts of ego-centric navigation data that is freely available on the internet can advance generalized intelligent systems, i.e., to robustly scale across perspectives, platforms, environmental conditions, scenarios, and geographical locations. However, it is difficult to directly leverage such large amounts of unlabeled and highly diverse data for complex 3D reasoning and planning tasks. Consequently, researchers have primarily focused on its use for various auxiliary pixel- and image-level computer vision tasks that do not consider an ultimate navigational objective. In this work, we introduce SelfD, a framework for learning scalable driving by utilizing large amounts of online monocular images. Our key idea is to leverage iterative semi-supervised training when learning imitative agents from unlabeled data. To handle unconstrained viewpoints, scenes, and camera parameters, we train an image-based model that directly learns to plan in the Bird's Eye View (BEV) space. Next, we use unlabeled data to augment the decision-making knowledge and robustness of an initially trained model via self-training. In particular, we propose a pseudo-labeling step which enables making full use of highly diverse demonstration data through "hypothetical" planning-based data augmentation. We employ a large dataset of publicly available YouTube videos to train SelfD and comprehensively analyze its generalization benefits across challenging navigation scenarios. Without requiring any additional data collection or annotation efforts, SelfD demonstrates consistent improvements (by up to 24%) in driving performance evaluation on nuScenes, Argoverse, Waymo, and CARLA.

Abstract— At present, inspection systems process visual data captured by cameras, with deep learning approaches applied to detect defects. Defect detection results usually have an accuracy higher than 94%. Real-life applications, however, are not very common. In this paper, we describe the development of a tire inspection system for the tire industry. We provide methods for processing tire sidewall data obtained from a camera and a laser sensor. The captured data comprise visual and geometric data characterizing the tire surface, providing a real representation of the captured tire sidewall. We use an unfolding process, that is, a polar transform, to further process the camera-obtained data. The principles and automation of the designed polar transform, based on polynomial regression (i.e., supervised learning), are presented. Based on the data from the laser sensor, the detection of abnormalities is performed using an unsupervised clustering method, followed by the classification of defects using the VGG-16 neural network. The inspection system aims to detect trained and untrained abnormalities, namely defects, as opposed to using only supervised learning methods. View Full-Text

Abstract— Camera calibration is integral to robotics and computer vision algorithms that seek to infer geometric properties of the scene from visual input streams. In practice, calibration is a laborious procedure requiring specialized data collection and careful tuning. This process must be repeated whenever the parameters of the camera change, which can be a frequent occurrence for mobile robots and autonomous vehicles. In contrast, self-supervised depth and ego-motion estimation approaches can bypass explicit calibration by inferring per-frame projection models that optimize a view synthesis objective. In this paper, we extend this approach to explicitly calibrate a wide range of cameras from raw videos in the wild. We propose a learning algorithm to regress per-sequence calibration parameters using an efficient family of general camera models. Our procedure achieves self-calibration results with sub-pixel reprojection error, outperforming other learning-based methods. We validate our approach on a wide variety of camera geometries, including perspective, fisheye, and catadioptric. Finally, we show that our approach leads to improvements in the downstream task of depth estimation, achieving state-of-the-art results on the EuRoC dataset with greater computational efficiency than contemporary methods.

Abstract— We propose an unsupervised, mid-level representation for a generative model of scenes. The representation is mid-level in that it is neither per-pixel nor per-image; rather, scenes are modeled as a collection of spatial, depth-ordered "blobs" of features. Blobs are differentiably placed onto a feature grid that is decoded into an image by a generative adversarial network. Due to the spatial uniformity of blobs and the locality inherent to convolution, our network learns to associate different blobs with different entities in a scene and to arrange these blobs to capture scene layout. We demonstrate this emergent behavior by showing that, despite training without any supervision, our method enables applications such as easy manipulation of objects within a scene (e.g., moving, removing, and restyling furniture), creation of feasible scenes given constraints (e.g., plausible rooms with drawers at a particular location), and parsing of real-world images into constituent parts. On a challenging multi-category dataset of indoor scenes, BlobGAN outperforms StyleGAN2 in image quality as measured by FID. See our project page for video results and interactive demo: this https URL.

Abstract— Progress in 3D object understanding has relied on manually canonicalized shape datasets that contain instances with consistent position and orientation (3D pose). This has made it hard to generalize these methods to in-the-wild shapes, eg., from internet model collections or depth sensors. ConDor is a self-supervised method that learns to Canonicalize the 3D orientation and position for full and partial 3D point clouds. We build on top of Tensor Field Networks (TFNs), a class of permutation- and rotation-equivariant, and translation-invariant 3D networks. During inference, our method takes an unseen full or partial 3D point cloud at an arbitrary pose and outputs an equivariant canonical pose. During training, this network uses self-supervision losses to learn the canonical pose from an un-canonicalized collection of full and partial 3D point clouds. ConDor can also learn to consistently co-segment object parts without any supervision. Extensive quantitative results on four new metrics show that our approach outperforms existing methods while enabling new applications such as operation on depth images and annotation transfer.

Abstract— Most natural videos contain numerous events. For example, in a video of a "man playing a piano", the video might also contain "another man dancing" or "a crowd clapping". We introduce the task of dense-captioning events, which involves both detecting and describing events in a video. We propose a new model that is able to identify all events in a single pass of the video while simultaneously describing the detected events with natural language. Our model introduces a variant of an existing proposal module that is designed to capture both short as well as long events that span minutes. To capture the dependencies between the events in a video, our model introduces a new captioning module that uses contextual information from past and future events to jointly describe all events. We also introduce ActivityNet Captions, a large-scale benchmark for dense-captioning events. ActivityNet Captions contains 20k videos amounting to 849 video hours with 100k total descriptions, each with it's unique start and end time. Finally, we report performances of our model for dense-captioning events, video retrieval and localization.

Abstract— We, humans, are entering into a virtual era and indeed want to bring animals to the virtual world as well for companion. Yet, computer-generated (CGI) furry animals are limited by tedious off-line rendering, let alone interactive motion control. In this paper, we present ARTEMIS, a novel neural modeling and rendering pipeline for generating ARTiculated neural pets with appEarance and Motion synthesIS. Our ARTEMIS enables interactive motion control, real-time animation, and photo-realistic rendering of furry animals. The core of our ARTEMIS is a neural-generated (NGI) animal engine, which adopts an efficient octree-based representation for animal animation and fur rendering. The animation then becomes equivalent to voxel-level deformation based on explicit skeletal warping. We further use a fast octree indexing and efficient volumetric rendering scheme to generate appearance and density features maps. Finally, we propose a novel shading network to generate high-fidelity details of appearance and opacity under novel poses from appearance and density feature maps. For the motion control module in ARTEMIS, we combine state-of-the-art animal motion capture approach with recent neural character control scheme. We introduce an effective optimization scheme to reconstruct the skeletal motion of real animals captured by a multi-view RGB and Vicon camera array. We feed all the captured motion into a neural character control scheme to generate abstract control signals with motion styles. We further integrate ARTEMIS into existing engines that support VR headsets, providing an unprecedented immersive experience where a user can intimately interact with a variety of virtual animals with vivid movements and photo-realistic appearance. We make available our ARTEMIS model and dynamic furry animal dataset at this https URL.

Abstract— We consider the targeted image editing problem: blending a region in a source image with a driver image that specifies the desired change. Differently from prior works, we solve this problem by learning a conditional probability distribution of the edits, end-to-end. Training such a model requires addressing a fundamental technical challenge: the lack of example edits for training. To this end, we propose a self-supervised approach that simulates edits by augmenting off-the-shelf images in a target domain. The benefits are remarkable: implemented as a state-of-the-art auto-regressive transformer, our approach is simple, sidesteps difficulties with previous methods based on GAN-like priors, obtains significantly better edits, and is efficient. Furthermore, we show that different blending effects can be learned by an intuitive control of the augmentation process, with no other changes required to the model architecture. We demonstrate the superiority of this approach across several datasets in extensive quantitative and qualitative experiments, including human studies, significantly outperforming prior work.

Abstract— The advent of large-scale training has produced a cornucopia of powerful visual recognition models. However, generative models, such as GANs, have traditionally been trained from scratch in an unsupervised manner. Can the collective "knowledge" from a large bank of pretrained vision models be leveraged to improve GAN training? If so, with so many models to choose from, which one(s) should be selected, and in what manner are they most effective? We find that pretrained computer vision models can significantly improve performance when used in an ensemble of discriminators. Notably, the particular subset of selected models greatly affects performance. We propose an effective selection mechanism, by probing the linear separability between real and fake samples in pretrained model embeddings, choosing the most accurate model, and progressively adding it to the discriminator ensemble. Interestingly, our method can improve GAN training in both limited data and large-scale settings. Given only 10k training samples, our FID on LSUN Cat matches the StyleGAN2 trained on 1.6M images. On the full dataset, our method improves FID by 1.5x to 2x on cat, church, and horse categories of LSUN.

Abstract— We present a new method that views object detection as a direct set prediction problem. Our approach streamlines the detection pipeline, effectively removing the need for many hand-designed components like a non-maximum suppression procedure or anchor generation that explicitly encode our prior knowledge about the task. The main ingredients of the new framework, called DEtection TRansformer or DETR, are a set-based global loss that forces unique predictions via bipartite matching, and a transformer encoder-decoder architecture. Given a fixed small set of learned object queries, DETR reasons about the relations of the objects and the global image context to directly output the final set of predictions in parallel. The new model is conceptually simple and does not require a specialized library, unlike many other modern detectors. DETR demonstrates accuracy and run-time performance on par with the well-established and highly-optimized Faster RCNN baseline on the challenging COCO object detection dataset. Moreover, DETR can be easily generalized to produce panoptic segmentation in a unified manner. We show that it significantly outperforms competitive baselines. Training code and pretrained models are available at this https URL.

Abstract— Inspired by progress in large-scale language modelling, we apply a similar approach towards building a single generalist agent beyond the realm of text outputs. The agent, which we refer to as Gato, works as a multi-modal, multi-task, multi-embodiment generalist policy. The same network with the same weights can play Atari, caption images, chat, stack blocks with a real robot arm and much more, deciding based on its context whether to output text, joint torques, button presses, or other tokens. In this report we describe the model and the data, and document the current capabilities of Gato.

Abstract— Physically-based differentiable rendering has recently emerged as an attractive new technique for solving inverse problems that recover complete 3D scene representations from images. The inversion of shape parameters is of particular interest but also poses severe challenges: shapes are intertwined with visibility, whose discontinuous nature introduces severe bias in computed derivatives unless costly precautions are taken. Shape representations like triangle meshes suffer from additional difficulties, since the continuous optimization of mesh parameters cannot introduce topological changes. One common solution to these difficulties entails representing shapes using signed distance functions (SDFs) and gradually adapting their zero level set during optimization. Previous differentiable rendering of SDFs did not fully account for visibility gradients and required the use of mask or silhouette supervision, or discretization into a triangle mesh. In this article, we show how to extend the commonly used sphere tracing algorithm so that it additionally outputs a reparameterization that provides the means to compute accurate shape parameter derivatives. At a high level, this resembles techniques for differentiable mesh rendering, though we show that the SDF representation admits a particularly efficient reparameterization that outperforms prior work. Our experiments demonstrate the reconstruction of (synthetic) objects without complex regularization or priors, using only a per-pixel RGB loss.

Abstract— Existing self-supervised 3D human pose estimation schemes have largely relied on weak supervisions like consistency loss to guide the learning, which, inevitably, leads to inferior results in real-world scenarios with unseen poses. In this paper, we propose a novel self-supervised approach that allows us to explicitly generate 2D-3D pose pairs for augmenting supervision, through a self-enhancing dual-loop learning framework. This is made possible via introducing a reinforcement-learning-based imitator, which is learned jointly with a pose estimator alongside a pose hallucinator; the three components form two loops during the training process, complementing and strengthening one another. Specifically, the pose estimator transforms an input 2D pose sequence to a low-fidelity 3D output, which is then enhanced by the imitator that enforces physical constraints. The refined 3D poses are subsequently fed to the hallucinator for producing even more diverse data, which are, in turn, strengthened by the imitator and further utilized to train the pose estimator. Such a co-evolution scheme, in practice, enables training a pose estimator on self-generated motion data without relying on any given 3D data. Extensive experiments across various benchmarks demonstrate that our approach yields encouraging results significantly outperforming the state of the art and, in some cases, even on par with results of fully-supervised methods. Notably, it achieves 89.1% 3D PCK on MPI-INF-3DHP under self-supervised cross-dataset evaluation setup, improving upon the previous best self-supervised methods by 8.6%. Code can be found at: this https URL.

Abstract— We propose a novel optimization-based paradigm for 3D human model fitting on images and scans. In contrast to existing approaches that directly regress the parameters of a low-dimensional statistical body model (e.g. SMPL) from input images, we train an ensemble of per-vertex neural fields network. The network predicts, in a distributed manner, the vertex descent direction towards the ground truth, based on neural features extracted at the current vertex projection. At inference, we employ this network, dubbed LVD, within a gradient-descent optimization pipeline until its convergence, which typically occurs in a fraction of a second even when initializing all vertices into a single point. An exhaustive evaluation demonstrates that our approach is able to capture the underlying body of clothed people with very different body shapes, achieving a significant improvement compared to state-of-the-art. LVD is also applicable to 3D model fitting of humans and hands, for which we show a significant improvement to the SOTA with a much simpler and faster method.

Abstract— We present Imagen, a text-to-image diffusion model with an unprecedented degree of photorealism and a deep level of language understanding. Imagen builds on the power of large transformer language models in understanding text and hinges on the strength of diffusion models in high-fidelity image generation. Our key discovery is that generic large language models (e.g. T5), pretrained on text-only corpora, are surprisingly effective at encoding text for image synthesis: increasing the size of the language model in Imagen boosts both sample fidelity and image-text alignment much more than increasing the size of the image diffusion model. Imagen achieves a new state-of-the-art FID score of 7.27 on the COCO dataset, without ever training on COCO, and human raters find Imagen samples to be on par with the COCO data itself in image-text alignment. To assess text-to-image models in greater depth, we introduce DrawBench, a comprehensive and challenging benchmark for text-to-image models. With DrawBench, we compare Imagen with recent methods including VQ-GAN+CLIP, Latent Diffusion Models, and DALL-E 2, and find that human raters prefer Imagen over other models in side-by-side comparisons, both in terms of sample quality and image-text alignment. See https://imagen.research.google/ for an overview of the results.

Abstract— Deep neural networks are vulnerable to attacks from adversarial inputs and, more recently, Trojans to misguide or hijack the model's decision. We expose the existence of an intriguing class of spatially bounded, physically realizable, adversarial examples -- Universal NaTuralistic adversarial paTches -- we call TnTs, by exploring the superset of the spatially bounded adversarial example space and the natural input space within generative adversarial networks. Now, an adversary can arm themselves with a patch that is naturalistic, less malicious-looking, physically realizable, highly effective achieving high attack success rates, and universal. A TnT is universal because any input image captured with a TnT in the scene will: i) misguide a network (untargeted attack); or ii) force the network to make a malicious decision (targeted attack). Interestingly, now, an adversarial patch attacker has the potential to exert a greater level of control -- the ability to choose a location-independent, natural-looking patch as a trigger in contrast to being constrained to noisy perturbations -- an ability is thus far shown to be only possible with Trojan attack methods needing to interfere with the model building processes to embed a backdoor at the risk discovery; but, still realize a patch deployable in the physical world. Through extensive experiments on the large-scale visual classification task, ImageNet with evaluations across its entire validation set of 50,000 images, we demonstrate the realistic threat from TnTs and the robustness of the attack. We show a generalization of the attack to create patches achieving higher attack success rates than existing state-of-the-art methods. Our results show the generalizability of the attack to different visual classification tasks (CIFAR-10, GTSRB, PubFig) and multiple state-of-the-art deep neural networks such as WideResnet50, Inception-V3 and VGG-16.

Abstract— We attempt to reduce the computational costs in vision transformers (ViTs), which increase quadratically in the token number. We present a novel training paradigm that trains only one ViT model at a time, but is capable of providing improved image recognition performance with various computational costs. Here, the trained ViT model, termed super vision transformer (SuperViT), is empowered with the versatile ability to solve incoming patches of multiple sizes as well as preserve informative tokens with multiple keeping rates (the ratio of keeping tokens) to achieve good hardware efficiency for inference, given that the available hardware resources often change from time to time. Experimental results on ImageNet demonstrate that our SuperViT can considerably reduce the computational costs of ViT models with even performance increase. For example, we reduce 2x FLOPs of DeiT-S while increasing the Top-1 accuracy by 0.2% and 0.7% for 1.5x reduction. Also, our SuperViT significantly outperforms existing studies on efficient vision transformers. For example, when consuming the same amount of FLOPs, our SuperViT surpasses the recent state-of-the-art (SoTA) EViT by 1.1% when using DeiT-S as their backbones. The project of this work is made publicly available at this https URL.

Abstract— Unsupervised object-centric learning aims to represent the modular, compositional, and causal structure of a scene as a set of object representations and thereby promises to resolve many critical limitations of traditional single-vector representations such as poor systematic generalization. Although there have been many remarkable advances in recent years, one of the most critical problems in this direction has been that previous methods work only with simple and synthetic scenes but not with complex and naturalistic images or videos. In this paper, we propose STEVE, an unsupervised model for object-centric learning in videos. Our proposed model makes a significant advancement by demonstrating its effectiveness on various complex and naturalistic videos unprecedented in this line of research. Interestingly, this is achieved by neither adding complexity to the model architecture nor introducing a new objective or weak supervision. Rather, it is achieved by a surprisingly simple architecture that uses a transformer-based image decoder conditioned on slots and the learning objective is simply to reconstruct the observation. Our experiment results on various complex and naturalistic videos show significant improvements compared to the previous state-of-the-art.

Abstract— Masked image modeling (MIM) learns representations with remarkably good fine-tuning performances, overshadowing previous prevalent pre-training approaches such as image classification, instance contrastive learning, and image-text alignment. In this paper, we show that the inferior fine-tuning performance of these pre-training approaches can be significantly improved by a simple post-processing in the form of feature distillation (FD). The feature distillation converts the old representations to new representations that have a few desirable properties just like those representations produced by MIM. These properties, which we aggregately refer to as optimization friendliness, are identified and analyzed by a set of attention- and optimization-related diagnosis tools. With these properties, the new representations show strong fine-tuning performance. Specifically, the contrastive self-supervised learning methods are made as competitive in fine-tuning as the state-of-the-art masked image modeling (MIM) algorithms. The CLIP models' fine-tuning performance is also significantly improved, with a CLIP ViT-L model reaching 89.0% top-1 accuracy on ImageNet-1K classification. On the 3-billion-parameter SwinV2-G model, the fine-tuning accuracy is improved by +1.5 mIoU / +1.1 mAP to 61.4 mIoU / 64.2 mAP on ADE20K semantic segmentation and COCO object detection, respectively, creating new records on both benchmarks. More importantly, our work provides a way for the future research to focus more effort on the generality and scalability of the learnt representations without being pre-occupied with optimization friendliness since it can be enhanced rather easily. The code will be available at this https URL.

Abstract— Deep networks are prone to performance degradation when there is a domain shift between the source (training) data and target (test) data. Recent test-time adaptation methods update batch normalization layers of pre-trained source models deployed in new target environments with streaming data to mitigate such performance degradation. Although such methods can adapt on-the-fly without first collecting a large target domain dataset, their performance is dependent on streaming conditions such as mini-batch size and class-distribution, which can be unpredictable in practice. In this work, we propose a framework for few-shot domain adaptation to address the practical challenges of data-efficient adaptation. Specifically, we propose a constrained optimization of feature normalization statistics in pre-trained source models supervised by a small support set from the target domain. Our method is easy to implement and improves source model performance with as few as one sample per class for classification tasks. Extensive experiments on 5 cross-domain classification and 4 semantic segmentation datasets show that our method achieves more accurate and reliable performance than test-time adaptation, while not being constrained by streaming conditions.

Abstract— This paper revisits visual representation in knowledge-based visual question answering (VQA) and demonstrates that using regional information in a better way can significantly improve the performance. While visual representation is extensively studied in traditional VQA, it is under-explored in knowledge-based VQA even though these two tasks share the common spirit, i.e., rely on visual input to answer the question. Specifically, we observe that in most state-of-the-art knowledge-based VQA methods: 1) visual features are extracted either from the whole image or in a sliding window manner for retrieving knowledge, and the important relationship within/among object regions is neglected; 2) visual features are not well utilized in the final answering model, which is counter-intuitive to some extent. Based on these observations, we propose a new knowledge-based VQA method REVIVE, which tries to utilize the explicit information of object regions not only in the knowledge retrieval stage but also in the answering model. The key motivation is that object regions and inherent relationships are important for knowledge-based VQA. We perform extensive experiments on the standard OK-VQA dataset and achieve new state-of-the-art performance, i.e., 58.0 state-of-the-art method by a large margin (+3.6 analysis and show the necessity of regional information in different framework components for knowledge-based VQA.

Abstract— Inverse rendering of an object under entirely unknown capture conditions is a fundamental challenge in computer vision and graphics. Neural approaches such as NeRF have achieved photorealistic results on novel view synthesis, but they require known camera poses. Solving this problem with unknown camera poses is highly challenging as it requires joint optimization over shape, radiance, and pose. This problem is exacerbated when the input images are captured in the wild with varying backgrounds and illuminations. Standard pose estimation techniques fail in such image collections in the wild due to very few estimated correspondences across images. Furthermore, NeRF cannot relight a scene under any illumination, as it operates on radiance (the product of reflectance and illumination). We propose a joint optimization framework to estimate the shape, BRDF, and per-image camera pose and illumination. Our method works on in-the-wild online image collections of an object and produces relightable 3D assets for several use-cases such as AR/VR. To our knowledge, our method is the first to tackle this severely unconstrained task with minimal user interaction. Project page: this https URL Video: this https URL

Abstract— Emerging neural radiance fields (NeRF) are a promising scene representation for computer graphics, enabling high-quality 3D reconstruction and novel view synthesis from image observations. However, editing a scene represented by a NeRF is challenging, as the underlying connectionist representations such as MLPs or voxel grids are not object-centric or compositional. In particular, it has been difficult to selectively edit specific regions or objects. In this work, we tackle the problem of semantic scene decomposition of NeRFs to enable query-based local editing of the represented 3D scenes. We propose to distill the knowledge of off-the-shelf, self-supervised 2D image feature extractors such as CLIP-LSeg or DINO into a 3D feature field optimized in parallel to the radiance field. Given a user-specified query of various modalities such as text, an image patch, or a point-and-click selection, 3D feature fields semantically decompose 3D space without the need for re-training and enable us to semantically select and edit regions in the radiance field. Our experiments validate that the distilled feature fields (DFFs) can transfer recent progress in 2D vision and language foundation models to 3D scene representations, enabling convincing 3D segmentation and selective editing of emerging neural graphics representations.

Abstract— We present a large-scale synthetic dataset for novel view synthesis consisting of ~300k images rendered from nearly 2000 complex scenes using high-quality ray tracing at high resolution (1600 x 1600 pixels). The dataset is orders of magnitude larger than existing synthetic datasets for novel view synthesis, thus providing a large unified benchmark for both training and evaluation. Using 4 distinct sources of high-quality 3D meshes, the scenes of our dataset exhibit challenging variations in camera views, lighting, shape, materials, and textures. Because our dataset is too large for existing methods to process, we propose Sparse Voxel Light Field (SVLF), an efficient voxel-based light field approach for novel view synthesis that achieves comparable performance to NeRF on synthetic data, while being an order of magnitude faster to train and two orders of magnitude faster to render. SVLF achieves this speed by relying on a sparse voxel octree, careful voxel sampling (requiring only a handful of queries per ray), and reduced network structure; as well as ground truth depth maps at training time. Our dataset is generated by NViSII, a Python-based ray tracing renderer, which is designed to be simple for non-experts to use and share, flexible and powerful through its use of scripting, and able to create high-quality and physically-based rendered images. Experiments with a subset of our dataset allow us to compare standard methods like NeRF and mip-NeRF for single-scene modeling, and pixelNeRF for category-level modeling, pointing toward the need for future improvements in this area.

Abstract— We address the task of view synthesis, which can be posed as recovering a rendering function that renders new views from a set of existing images. In many recent works such as NeRF, this rendering function is parameterized using implicit neural representations of scene geometry. Implicit neural representations have achieved impressive visual quality but have drawbacks in computational efficiency. In this work, we propose a new approach that performs view synthesis using point clouds. It is the first point-based method to achieve better visual quality than NeRF while being more than 100x faster in rendering speed. Our approach builds on existing works on differentiable point-based rendering but introduces a novel technique we call "Sculpted Neural Points (SNP)", which significantly improves the robustness to errors and holes in the reconstructed point cloud. Experiments show that on the task of view synthesis, our sculpting technique closes the gap between point-based and implicit representation-based methods. Code is available at this https URL and supplementary video at this https URL.

Abstract— We present a phasorial embedding field \emph{PREF} as a compact representation to facilitate neural signal modeling and reconstruction tasks. Pure multi-layer perceptron (MLP) based neural techniques are biased towards low frequency signals and have relied on deep layers or Fourier encoding to avoid losing details. PREF instead employs a compact and physically explainable encoding field based on the phasor formulation of the Fourier embedding space. We conduct comprehensive experiments to demonstrate the advantages of PREF over the latest spatial embedding techniques. We then develop a highly efficient frequency learning framework using an approximated inverse Fourier transform scheme for PREF along with a novel Parseval regularizer. Extensive experiments show our efficient and compact frequency-based neural signal processing technique is on par with and even better than the state-of-the-art in 2D image completion, 3D SDF surface regression, and 5D radiance field reconstruction.

Abstract— We present a video generation model that accurately reproduces object motion, changes in camera viewpoint, and new content that arises over time. Existing video generation methods often fail to produce new content as a function of time while maintaining consistencies expected in real environments, such as plausible dynamics and object persistence. A common failure case is for content to never change due to over-reliance on inductive biases to provide temporal consistency, such as a single latent code that dictates content for the entire video. On the other extreme, without long-term consistency, generated videos may morph unrealistically between different scenes. To address these limitations, we prioritize the time axis by redesigning the temporal latent representation and learning long-term consistency from data by training on longer videos. To this end, we leverage a two-phase training strategy, where we separately train using longer videos at a low resolution and shorter videos at a high resolution. To evaluate the capabilities of our model, we introduce two new benchmark datasets with explicit focus on long-term temporal dynamics.

Abstract— Employing deep learning-based approaches for fine-grained facial expression analysis, such as those involving the estimation of Action Unit (AU) intensities, is difficult due to the lack of a large-scale dataset of real faces with sufficiently diverse AU labels for training. In this paper, we consider how AU-level facial image synthesis can be used to substantially augment such a dataset. We propose an AU synthesis framework that combines the well-known 3D Morphable Model (3DMM), which intrinsically disentangles expression parameters from other face attributes, with models that adversarially generate 3DMM expression parameters conditioned on given target AU labels, in contrast to the more conventional approach of generating facial images directly. In this way, we are able to synthesize new combinations of expression parameters and facial images from desired AU labels. Extensive quantitative and qualitative results on the benchmark DISFA dataset demonstrate the effectiveness of our method on 3DMM facial expression parameter synthesis and data augmentation for deep learning-based AU intensity estimation.

Abstract— Unsupervised generation of high-quality multi-view-consistent images and 3D shapes using only collections of single-view 2D photographs has been a long-standing challenge. Existing 3D GANs are either compute-intensive or make approximations that are not 3D-consistent; the former limits quality and resolution of the generated images and the latter adversely affects multi-view consistency and shape quality. In this work, we improve the computational efficiency and image quality of 3D GANs without overly relying on these approximations. We introduce an expressive hybrid explicit-implicit network architecture that, together with other design choices, synthesizes not only high-resolution multi-view-consistent images in real time but also produces high-quality 3D geometry. By decoupling feature generation and neural rendering, our framework is able to leverage state-of-the-art 2D CNN generators, such as StyleGAN2, and inherit their efficiency and expressiveness. We demonstrate state-of-the-art 3D-aware synthesis with FFHQ and AFHQ Cats, among other experiments.

Abstract— Efficient neural network backbones for mobile devices are often optimized for metrics such as FLOPs or parameter count. However, these metrics may not correlate well with latency of the network when deployed on a mobile device. Therefore, we perform extensive analysis of different metrics by deploying several mobile-friendly networks on a mobile device. We identify and analyze architectural and optimization bottlenecks in recent efficient neural networks and provide ways to mitigate these bottlenecks. To this end, we design an efficient backbone MobileOne, with variants achieving an inference time under 1 ms on an iPhone12 with 75.9% top-1 accuracy on ImageNet. We show that MobileOne achieves state-of-the-art performance within the efficient architectures while being many times faster on mobile. Our best model obtains similar performance on ImageNet as MobileFormer while being 38x faster. Our model obtains 2.3% better top-1 accuracy on ImageNet than EfficientNet at similar latency. Furthermore, we show that our model generalizes to multiple tasks - image classification, object detection, and semantic segmentation with significant improvements in latency and accuracy as compared to existing efficient architectures when deployed on a mobile device.

Abstract— Positional encoding is important for vision transformer (ViT) to capture the spatial structure of the input image. General effectiveness has been proven in ViT. In our work we propose to train ViT to recognize the positional label of patches of the input image, this apparently simple task actually yields a meaningful self-supervisory task. Based on previous work on ViT positional encoding, we propose two positional labels dedicated to 2D images including absolute position and relative position. Our positional labels can be easily plugged into various current ViT variants. It can work in two ways: (a) As an auxiliary training target for vanilla ViT (e.g., ViT-B and Swin-B) for better performance. (b) Combine the self-supervised ViT (e.g., MAE) to provide a more powerful self-supervised signal for semantic feature learning. Experiments demonstrate that with the proposed self-supervised methods, ViT-B and Swin-B gain improvements of 1.20% (top-1 Acc) and 0.74% (top-1 Acc) on ImageNet, respectively, and 6.15% and 1.14% improvement on Mini-ImageNet.

Abstract— Unsupervised representation learning methods like SwAV are proved to be effective in learning visual semantics of a target dataset. The main idea behind these methods is that different views of a same image represent the same semantics. In this paper, we further introduce an add-on module to facilitate the injection of the knowledge accounting for spatial cross correlations among the samples. This in turn results in distilling intra-class information including feature level locations and cross similarities between same-class instances. The proposed add-on can be added to existing methods such as the SwAV. We can later remove the add-on module for inference without any modification of the learned weights. Through an extensive set of empirical evaluations, we verify that our method yields an improved performance in detecting the class activation maps, top-1 classification accuracy, and down-stream tasks such as object detection, with different configuration settings.

Abstract— Image synthesis under multi-modal priors is a useful and challenging task that has received increasing attention in recent years. A major challenge in using generative models to accomplish this task is the lack of paired data containing all modalities (i.e. priors) and corresponding outputs. In recent work, a variational auto-encoder (VAE) model was trained in a weakly supervised manner to address this challenge. Since the generative power of VAEs is usually limited, it is difficult for this method to synthesize images belonging to complex distributions. To this end, we propose a solution based on a denoising diffusion probabilistic models to synthesise images under multi-model priors. Based on the fact that the distribution over each time step in the diffusion model is Gaussian, in this work we show that there exists a closed-form expression to the generate the image corresponds to the given modalities. The proposed solution does not require explicit retraining for all modalities and can leverage the outputs of individual modalities to generate realistic images according to different constraints. We conduct studies on two real-world datasets to demonstrate the effectiveness of our approach.

Abstract— From CNN, RNN, to ViT, we have witnessed remarkable advancements in video prediction, incorporating auxiliary inputs, elaborate neural architectures, and sophisticated training strategies. We admire these progresses but are confused about the necessity: is there a simple method that can perform comparably well? This paper proposes SimVP, a simple video prediction model that is completely built upon CNN and trained by MSE loss in an end-to-end fashion. Without introducing any additional tricks and complicated strategies, we can achieve state-of-the-art performance on five benchmark datasets. Through extended experiments, we demonstrate that SimVP has strong generalization and extensibility on real-world datasets. The significant reduction of training cost makes it easier to scale to complex scenarios. We believe SimVP can serve as a solid baseline to stimulate the further development of video prediction. The code is available at this https URL.

Abstract— We examine how the saccade mechanism from biological vision can be used to make deep neural networks more efficient for classification and object detection problems. Our proposed approach is based on the ideas of attention-driven visual processing and saccades, miniature eye movements influenced by attention. We conduct experiments by analyzing: i) the robustness of different deep neural network (DNN) feature extractors to partially-sensed images for image classification and object detection, and ii) the utility of saccades in masking image patches for image classification and object tracking. Experiments with convolutional nets (ResNet-18) and transformer-based models (ViT, DETR, TransTrack) are conducted on several datasets (CIFAR-10, DAVSOD, MSCOCO, and MOT17). Our experiments show intelligent data reduction via learning to mimic human saccades when used in conjunction with state-of-the-art DNNs for classification, detection, and tracking tasks. We observed minimal drop in performance for the classification and detection tasks while only using about 30% of the original sensor data. We discuss how the saccade mechanism can inform hardware design via ``in-pixel'' processing.

Abstract— Event cameras are bio-inspired sensors that perform well in challenging illumination conditions and have high temporal resolution. However, their concept is fundamentally different from traditional frame-based cameras. The pixels of an event camera operate independently and asynchronously. They measure changes of the logarithmic brightness and return them in the highly discretised form of time-stamped events indicating a relative change of a certain quantity since the last event. New models and algorithms are needed to process this kind of measurements. The present work looks at several motion estimation problems with event cameras. The flow of the events is modelled by a general homographic warping in a space-time volume, and the objective is formulated as a maximisation of contrast within the image of warped events. Our core contribution consists of deriving globally optimal solutions to these generally non-convex problems, which removes the dependency on a good initial guess plaguing existing methods. Our methods rely on branch-and-bound optimisation and employ novel and efficient, recursive upper and lower bounds derived for six different contrast estimation functions. The practical validity of our approach is demonstrated by a successful application to three different event camera motion estimation problems.

Abstract— In this paper we present Hyper-Dimensional Reconfigurable Analytics at the Tactical Edge (HyDRATE) using low-SWaP embedded hardware that can perform real-time reconfiguration at the edge leveraging non-MAC (free of floating-point MultiplyACcumulate operations) deep neural nets (DNN) combined with hyperdimensional (HD) computing accelerators. We describe the algorithm, trained quantized model generation, and simulated performance of a feature extractor free of multiply-accumulates feeding a hyperdimensional logic-based classifier. Then we show how performance increases with the number of hyperdimensions. We describe the realized low-SWaP FPGA hardware and embedded software system compared to traditional DNNs and detail the implemented hardware accelerators. We discuss the measured system latency and power, noise robustness due to use of learnable quantization and HD computing, actual versus simulated system performance for a video activity classification task and demonstration of reconfiguration on this same dataset. We show that reconfigurability in the field is achieved by retraining only the feed-forward HD classifier without gradient descent backpropagation (gradient-free), using few-shot learning of new classes at the edge. Initial work performed used LRCN DNN and is currently extended to use Two-stream DNN with improved performance.

Abstract— Image matting refers to extracting the accurate foregrounds in the image. Current automatic methods tend to extract all the salient objects in the image indiscriminately. In this paper, we propose a new task named Referring Image Matting (RIM), referring to extracting the meticulous alpha matte of the specific object that can best match the given natural language description. However, prevalent visual grounding methods are all limited to the segmentation level, probably due to the lack of high-quality datasets for RIM. To fill the gap, we establish the first large-scale challenging dataset RefMatte by designing a comprehensive image composition and expression generation engine to produce synthetic images on top of current public high-quality matting foregrounds with flexible logics and re-labelled diverse attributes. RefMatte consists of 230 object categories, 47,500 images, 118,749 expression-region entities, and 474,996 expressions, which can be further extended easily in the future. Besides this, we also construct a real-world test set with manually generated phrase annotations consisting of 100 natural images to further evaluate the generalization of RIM models. We first define the task of RIM in two settings, i.e., prompt-based and expression-based, and then benchmark several representative methods together with specific model designs for image matting. The results provide empirical insights into the limitations of existing methods as well as possible solutions. We believe the new task RIM along with the RefMatte dataset will open new research directions in this area and facilitate future studies. The dataset and code will be made publicly available at this https URL.

Abstract— Advances in learning-based trajectory prediction are enabled by large-scale datasets. However, in-depth analysis of such datasets is limited. Moreover, the evaluation of prediction models is limited to metrics averaged over all samples in the dataset. We propose an automated methodology that allows to extract maneuvers (e.g., left turn, lane change) from agent trajectories in such datasets. The methodology considers information about the agent dynamics and information about the lane segments the agent traveled along. Although it is possible to use the resulting maneuvers for training classification networks, we exemplary use them for extensive trajectory dataset analysis and maneuver-specific evaluation of multiple state-of-the-art trajectory prediction models. Additionally, an analysis of the datasets and an evaluation of the prediction models based on the agent dynamics is provided.

Abstract— A new open-source image processing pipeline for analyzing camera trap time-lapse recordings is described. This pipeline includes machine learning models to assist human-in-the-loop video segmentation and animal re-identification. We present some performance results and observations on the utility of this pipeline after using it in a year-long project studying the spatial ecology and social behavior of the gopher tortoise.

Abstract— Learning representations with self-supervision for convolutional networks (CNN) has proven effective for vision tasks. As an alternative for CNN, vision transformers (ViTs) emerge strong representation ability with the pixel-level self-attention and channel-level feed-forward networks. Recent works reveal that self-supervised learning helps unleash the great potential of ViTs. Still, most works follow self-supervised strategy designed for CNNs, e.g., instance-level discrimination of samples, but they ignore the unique properties of ViTs. We observe that modeling relations among pixels and channels distinguishes ViTs from other networks. To enforce this property, we explore the feature self-relations for training self-supervised ViTs. Specifically, instead of conducting self-supervised learning solely on feature embeddings from multiple views, we utilize the feature self-relations, i.e., pixel/channel-level self-relations, for self-supervised learning. Self-relation based learning further enhance the relation modeling ability of ViTs, resulting in strong representations that stably improve performance on multiple downstream tasks. Our source code will be made publicly available.

Abstract— Embedding of large but redundant data, such as images or text, in a hierarchy of lower-dimensional spaces is one of the key features of representation learning approaches, which nowadays provide state-of-the-art solutions to problems once believed hard or impossible to solve. In this work, in a plot twist with a strong meta aftertaste, we show how trained deep models are as redundant as the data they are optimized to process, and how it is therefore possible to use deep learning models to embed deep learning models. In particular, we show that it is possible to use representation learning to learn a fixed-size, low-dimensional embedding space of trained deep models and that such space can be explored by interpolation or optimization to attain ready-to-use models. We find that it is possible to learn an embedding space of multiple instances of the same architecture and of multiple architectures. We address image classification and neural representation of signals, showing how our embedding space can be learnt so as to capture the notions of performance and 3D shape, respectively. In the Multi-Architecture setting we also show how an embedding trained only on a subset of architectures can learn to generate already-trained instances of architectures it never sees instantiated at training time.

Abstract— Unets have become the standard method for semantic segmentation of medical images, along with fully convolutional networks (FCN). Unet++ was introduced as a variant of Unet, in order to solve some of the problems facing Unet and FCNs. Unet++ provided networks with an ensemble of variable depth Unets, hence eliminating the need for professionals estimating the best suitable depth for a task. While Unet and all its variants, including Unet++ aimed at providing networks that were able to train well without requiring large quantities of annotated data, none of them attempted to eliminate the need for pixel-wise annotated data altogether. Obtaining such data for each disease to be diagnosed comes at a high cost. Hence such data is scarce. In this paper we use contrastive learning to train Unet++ for semantic segmentation of medical images using medical images from various sources including magnetic resonance imaging (MRI) and computed tomography (CT), without the need for pixel-wise annotations. Here we describe the architecture of the proposed model and the training method used. This is still a work in progress and so we abstain from including results in this paper. The results and the trained model would be made available upon publication or in subsequent versions of this paper on arxiv.

Abstract— We study how networking corruptions--data corruptions caused by networking errors--affect video machine learning (ML) models. We discover apparent networking corruptions in Kinetics-400, a benchmark video ML dataset. In a simulation study, we investigate (1) what artifacts networking corruptions cause, (2) how such artifacts affect ML models, and (3) whether standard robustness methods can mitigate their negative effects. We find that networking corruptions cause visual and temporal artifacts (i.e., smeared colors or frame drops). These networking corruptions degrade performance on a variety of video ML tasks, but effects vary by task and dataset, depending on how much temporal context the tasks require. Lastly, we evaluate data augmentation--a standard defense for data corruptions--but find that it does not recover performance.

Abstract— Previous work generally believes that improving the spatial invariance of convolutional networks is the key to object counting. However, after verifying several mainstream counting networks, we surprisingly found too strict pixel-level spatial invariance would cause overfit noise in the density map generation. In this paper, we try to use locally connected Gaussian kernels to replace the original convolution filter to estimate the spatial position in the density map. The purpose of this is to allow the feature extraction process to potentially stimulate the density map generation process to overcome the annotation noise. Inspired by previous work, we propose a low-rank approximation accompanied with translation invariance to favorably implement the approximation of massive Gaussian convolution. Our work points a new direction for follow-up research, which should investigate how to properly relax the overly strict pixel-level spatial invariance for object counting. We evaluate our methods on 4 mainstream object counting networks (i.e., MCNN, CSRNet, SANet, and ResNet-50). Extensive experiments were conducted on 7 popular benchmarks for 3 applications (i.e., crowd, vehicle, and plant counting). Experimental results show that our methods significantly outperform other state-of-the-art methods and achieve promising learning of the spatial position of objects.

Abstract— Despite their high accuracies, modern complex image classifiers cannot be trusted for sensitive tasks due to their unknown decision-making process and potential biases. Counterfactual explanations are very effective in providing transparency for these black-box algorithms. Nevertheless, generating counterfactuals that can have a consistent impact on classifier outputs and yet expose interpretable feature changes is a very challenging task. We introduce a novel method to generate causal and yet interpretable counterfactual explanations for image classifiers using pretrained generative models without any re-training or conditioning. The generative models in this technique are not bound to be trained on the same data as the target classifier. We use this framework to obtain contrastive and causal sufficiency and necessity scores as global explanations for black-box classifiers. On the task of face attribute classification, we show how different attributes influence the classifier output by providing both causal and contrastive feature attributions, and the corresponding counterfactual images.

Abstract— Neural radiance fields have made a remarkable breakthrough in the novel view synthesis task at the 3D static scene. However, for the 4D circumstance (e.g., dynamic scene), the performance of the existing method is still limited by the capacity of the neural network, typically in a multilayer perceptron network (MLP). In this paper, we present the method to model the 4D neural radiance field by the 3D voxel, short as V4D, where the 3D voxel has two formats. The first one is to regularly model the bounded 3D space and then use the sampled local 3D feature with the time index to model the density field and the texture field. The second one is in look-up tables (LUTs) format that is for the pixel-level refinement, where the pseudo-surface produced by the volume rendering is utilized as the guidance information to learn a 2D pixel-level refinement mapping. The proposed LUTs-based refinement module achieves the performance gain with a little computational cost and could serve as the plug-and-play module in the novel view synthesis task. Moreover, we propose a more effective conditional positional encoding toward the 4D data that achieves performance gain with negligible computational burdens. Extensive experiments demonstrate that the proposed method achieves state-of-the-art performance by a large margin. At last, the proposed V4D is also a computational-friendly method in both the training and testing phase, where we achieve 2 times faster in the training phase and 10 times faster in the inference phase compared with the state-of-the-art method.

Abstract— Neural Radiance Field (NeRF) has emerged as a compelling method to represent 3D objects and scenes for photo-realistic rendering. However, its implicit representation causes difficulty in manipulating the models like the explicit mesh representation. Several recent advances in NeRF manipulation are usually restricted by a shared renderer network, or suffer from large model size. To circumvent the hurdle, in this paper, we present an explicit neural field representation that enables efficient and convenient manipulation of models. To achieve this goal, we learn a hybrid tensor rank decomposition of the scene without neural networks. Motivated by the low-rank approximation property of the SVD algorithm, we propose a rank-residual learning strategy to encourage the preservation of primary information in lower ranks. The model size can then be dynamically adjusted by rank truncation to control the levels of detail, achieving near-optimal compression without extra optimization. Furthermore, different models can be arbitrarily transformed and composed into one scene by concatenating along the rank dimension. The growth of storage cost can also be mitigated by compressing the unimportant objects in the composed scene. We demonstrate that our method is able to achieve comparable rendering quality to state-of-the-art methods, while enabling extra capability of compression and composition. Code will be made available at this https URL.

Abstract— Rendering scenes with a high-quality human face from arbitrary viewpoints is a practical and useful technique for many real-world applications. Recently, Neural Radiance Fields (NeRF), a rendering technique that uses neural networks to approximate classical ray tracing, have been considered as one of the promising approaches for synthesizing novel views from a sparse set of images. We find that NeRF can render new views while maintaining geometric consistency, but it does not properly maintain skin details, such as moles and pores. These details are important particularly for faces because when we look at an image of a face, we are much more sensitive to details than when we look at other objects. On the other hand, 3D Morpable Models (3DMMs) based on traditional meshes and textures can perform well in terms of skin detail despite that it has less precise geometry and cannot cover the head and the entire scene with background. Based on these observations, we propose a method to use both NeRF and 3DMM to synthesize a high-fidelity novel view of a scene with a face. Our method learns a Generative Adversarial Network (GAN) to mix a NeRF-synthesized image and a 3DMM-rendered image and produces a photorealistic scene with a face preserving the skin details. Experiments with various real-world scenes demonstrate the effectiveness of our approach. The code will be available on this https URL.

Abstract— Neural radiance fields (NeRF) have shown great success in modeling 3D scenes and synthesizing novel-view images. However, most previous NeRF methods take much time to optimize one single scene. Explicit data structures, e.g. voxel features, show great potential to accelerate the training process. However, voxel features face two big challenges to be applied to dynamic scenes, i.e. modeling temporal information and capturing different scales of point motions. We propose a radiance field framework by representing scenes with time-aware voxel features, named as TiNeuVox. A tiny coordinate deformation network is introduced to model coarse motion trajectories and temporal information is further enhanced in the radiance network. A multi-distance interpolation method is proposed and applied on voxel features to model both small and large motions. Our framework significantly accelerates the optimization of dynamic radiance fields while maintaining high rendering quality. Empirical evaluation is performed on both synthetic and real scenes. Our TiNeuVox completes training with only 8 minutes and 8-MB storage cost while showing similar or even better rendering performance than previous dynamic NeRF methods.

Abstract— We introduce an approach for selecting objects in neural volumetric 3D representations, such as multi-plane images (MPI) and neural radiance fields (NeRF). Our approach takes a set of foreground and background 2D user scribbles in one view and automatically estimates a 3D segmentation of the desired object, which can be rendered into novel views. To achieve this result, we propose a novel voxel feature embedding that incorporates the neural volumetric 3D representation and multi-view image features from all input views. To evaluate our approach, we introduce a new dataset of human-provided segmentation masks for depicted objects in real-world multi-view scene captures. We show that our approach out-performs strong baselines, including 2D segmentation and 3D segmentation approaches adapted to our task.

Abstract— Given a monocular video, segmenting and decoupling dynamic objects while recovering the static environment is a widely studied problem in machine intelligence. Existing solutions usually approach this problem in the image domain, limiting their performance and understanding of the environment. We introduce Decoupled Dynamic Neural Radiance Field (D$^{2}$NeRF), a self-supervised approach that takes a monocular video and learns a 3D scene representation which decouples moving objects, including their shadows, from the static background. Our method represents the moving objects and the static background by two separate neural radiance fields with only one allowing for temporal changes. A naive implementation of this approach leads to the dynamic component taking over the static one as the representation of the former is inherently more general and prone to overfitting. To this end, we propose a novel loss to promote correct separation of phenomena. We further propose a shadow field network to detect and decouple dynamically moving shadows. We introduce a new dataset containing various dynamic objects and shadows and demonstrate that our method can achieve better performance than state-of-the-art approaches in decoupling dynamic and static 3D objects, occlusion and shadow removal, and image segmentation for moving objects.

Abstract— Neural Radiance Fields (NeRF) has been wildly applied to various tasks for its high-quality representation of 3D scenes. It takes long per-scene training time and per-image testing time. In this paper, we present EfficientNeRF as an efficient NeRF-based method to represent 3D scene and synthesize novel-view images. Although several ways exist to accelerate the training or testing process, it is still difficult to much reduce time for both phases simultaneously. We analyze the density and weight distribution of the sampled points then propose valid and pivotal sampling at the coarse and fine stage, respectively, to significantly improve sampling efficiency. In addition, we design a novel data structure to cache the whole scene during testing to accelerate the rendering speed. Overall, our method can reduce over 88% of training time, reach rendering speed of over 200 FPS, while still achieving competitive accuracy. Experiments prove that our method promotes the practicality of NeRF in the real world and enables many applications.

Abstract— Contemporary registration devices for 3D visual information, such as LIDARs and various depth cameras, capture data as 3D point clouds. In turn, such clouds are challenging to be processed due to their size and complexity. Existing methods address this problem by fitting a mesh to the point cloud and rendering it instead. This approach, however, leads to the reduced fidelity of the resulting visualization and misses color information of the objects crucial in computer graphics applications. In this work, we propose to mitigate this challenge by representing 3D objects as Neural Radiance Fields (NeRFs). We leverage a hypernetwork paradigm and train the model to take a 3D point cloud with the associated color values and return a NeRF network's weights that reconstruct 3D objects from input 2D images. Our method provides efficient 3D object representation and offers several advantages over the existing approaches, including the ability to condition NeRFs and improved generalization beyond objects seen in training. The latter we also confirmed in the results of our empirical evaluation.

Abstract— It is a long-standing problem to find effective representations for training reinforcement learning (RL) agents. This paper demonstrates that learning state representations with supervision from Neural Radiance Fields (NeRFs) can improve the performance of RL compared to other learned representations or even low-dimensional, hand-engineered state information. Specifically, we propose to train an encoder that maps multiple image observations to a latent space describing the objects in the scene. The decoder built from a latent-conditioned NeRF serves as the supervision signal to learn the latent space. An RL algorithm then operates on the learned latent space as its state representation. We call this NeRF-RL. Our experiments indicate that NeRF as supervision leads to a latent space better suited for the downstream RL tasks involving robotic object manipulations like hanging mugs on hooks, pushing objects, or opening doors. Video: this https URL.

Abstract— We present ObPose, an unsupervised object-centric generative model that learns to segment 3D objects from RGB-D video in an unsupervised manner. Inspired by prior art in 2D representation learning, ObPose considers a factorised latent space, separately encoding object-wise location (where) and appearance (what) information. In particular, ObPose leverages an object's canonical pose, defined via a minimum volume principle, as a novel inductive bias for learning the where component. To achieve this, we propose an efficient, voxelised approximation approach to recover the object shape directly from a neural radiance field (NeRF). As a consequence, ObPose models scenes as compositions of NeRFs representing individual objects. When evaluated on the YCB dataset for unsupervised scene segmentation, ObPose outperforms the current state-of-the-art in 3D scene inference (ObSuRF) by a significant margin in terms of segmentation quality for both video inputs as well as for multi-view static scenes. In addition, the design choices made in the ObPose encoder are validated with relevant ablations.

Abstract— Detecting 3D keypoints from point clouds is important for shape reconstruction, while this work investigates the dual question: can shape reconstruction benefit 3D keypoint detection? Existing methods either seek salient features according to statistics of different orders or learn to predict keypoints that are invariant to transformation. Nevertheless, the idea of incorporating shape reconstruction into 3D keypoint detection is under-explored. We argue that this is restricted by former problem formulations. To this end, a novel unsupervised paradigm named SNAKE is proposed, which is short for shape-aware neural 3D keypoint field. Similar to recent coordinate-based radiance or distance field, our network takes 3D coordinates as inputs and predicts implicit shape indicators and keypoint saliency simultaneously, thus naturally entangling 3D keypoint detection and shape reconstruction. We achieve superior performance on various public benchmarks, including standalone object datasets ModelNet40, KeypointNet, SMPL meshes and scene-level datasets 3DMatch and Redwood. Intrinsic shape awareness brings several advantages as follows. (1) SNAKE generates 3D keypoints consistent with human semantic annotation, even without such supervision. (2) SNAKE outperforms counterparts in terms of repeatability, especially when the input point clouds are down-sampled. (3) the generated keypoints allow accurate geometric registration, notably in a zero-shot setting. Codes are available at this https URL.

Abstract— Comprehensive 3D scene understanding, both geometrically and semantically, is important for real-world applications such as robot perception. Most of the existing work has focused on developing data-driven discriminative models for scene understanding. This paper provides a new approach to scene understanding, from a synthesis model perspective, by leveraging the recent progress on implicit 3D representation and neural rendering. Building upon the great success of Neural Radiance Fields (NeRFs), we introduce Scene-Property Synthesis with NeRF (SS-NeRF) that is able to not only render photo-realistic RGB images from novel viewpoints, but also render various accurate scene properties (e.g., appearance, geometry, and semantics). By doing so, we facilitate addressing a variety of scene understanding tasks under a unified framework, including semantic segmentation, surface normal estimation, reshading, keypoint detection, and edge detection. Our SS-NeRF framework can be a powerful tool for bridging generative learning and discriminative learning, and thus be beneficial to the investigation of a wide range of interesting problems, such as studying task relationships within a synthesis paradigm, transferring knowledge to novel tasks, facilitating downstream discriminative tasks as ways of data augmentation, and serving as auto-labeller for data creation.

Abstract— Face alignment, which fits a face model to an image and extracts the semantic meanings of facial pixels, has been an important topic in the computer vision community. However, most algorithms are designed for faces in small to medium poses (yaw angle is smaller than 45 degrees), which lack the ability to align faces in large poses up to 90 degrees. The challenges are three-fold. Firstly, the commonly used landmark face model assumes that all the landmarks are visible and is therefore not suitable for large poses. Secondly, the face appearance varies more drastically across large poses, from the frontal view to the profile view. Thirdly, labelling landmarks in large poses is extremely challenging since the invisible landmarks have to be guessed. In this paper, we propose to tackle these three challenges in an new alignment framework termed 3D Dense Face Alignment (3DDFA), in which a dense 3D Morphable Model (3DMM) is fitted to the image via Cascaded Convolutional Neural Networks. We also utilize 3D information to synthesize face images in profile views to provide abundant samples for training. Experiments on the challenging AFLW database show that the proposed approach achieves significant improvements over the state-of-the-art methods.

Abstract— We introduce a new neural signal model designed for efficient high-resolution representation of large-scale signals. The key innovation in our multiscale implicit neural representation (MINER) is an internal representation via a Laplacian pyramid, which provides a sparse multiscale decomposition of the signal that captures orthogonal parts of the signal across scales. We leverage the advantages of the Laplacian pyramid by representing small disjoint patches of the pyramid at each scale with a small MLP. This enables the capacity of the network to adaptively increase from coarse to fine scales, and only represent parts of the signal with strong signal energy. The parameters of each MLP are optimized from coarse-to-fine scale which results in faster approximations at coarser scales, thereby ultimately an extremely fast training process. We apply MINER to a range of large-scale signal representation tasks, including gigapixel images and very large point clouds, and demonstrate that it requires fewer than 25% of the parameters, 33% of the memory footprint, and 10% of the computation time of competing techniques such as ACORN to reach the same representation accuracy.

Abstract— Unsupervised generation of high-quality multi-view-consistent images and 3D shapes using only collections of single-view 2D photographs has been a long-standing challenge. Existing 3D GANs are either compute-intensive or make approximations that are not 3D-consistent; the former limits quality and resolution of the generated images and the latter adversely affects multi-view consistency and shape quality. In this work, we improve the computational efficiency and image quality of 3D GANs without overly relying on these approximations. We introduce an expressive hybrid explicit-implicit network architecture that, together with other design choices, synthesizes not only high-resolution multi-view-consistent images in real time but also produces high-quality 3D geometry. By decoupling feature generation and neural rendering, our framework is able to leverage state-of-the-art 2D CNN generators, such as StyleGAN2, and inherit their efficiency and expressiveness. We demonstrate state-of-the-art 3D-aware synthesis with FFHQ and AFHQ Cats, among other experiments.

Abstract— Lung cancer is the leading cause of cancer-related death worldwide. Lung adenocarcinoma (LUAD) and lung squamous cell carcinoma (LUSC) are the most common histologic subtypes of non-small-cell lung cancer (NSCLC). Histology is an essential tool for lung cancer diagnosis. Pathologists make classifications according to the dominant subtypes. Although morphology remains the standard for diagnosis, significant tool needs to be developed to elucidate the diagnosis. In our study, we utilize the pre-trained Vision Transformer (ViT) model to classify multiple label lung cancer on histologic slices (from dataset LC25000), in both Zero-Shot and Few-Shot settings. Then we compare the performance of Zero-Shot and Few-Shot ViT on accuracy, precision, recall, sensitivity and specificity. Our study show that the pre-trained ViT model has a good performance in Zero-Shot setting, a competitive accuracy (99.87%) in Few-Shot setting ({epoch = 1}) and an optimal result (100.00% on both validation set and test set) in Few-Shot seeting ({epoch = 5}).

Abstract— Capturing both local and global features of irregular point clouds is essential to 3D object detection (3OD). However, mainstream 3D detectors, e.g., VoteNet and its variants, either abandon considerable local features during pooling operations or ignore many global features in the whole scene context. This paper explores new modules to simultaneously learn local-global features of scene point clouds that serve 3OD positively. To this end, we propose an effective 3OD network via simultaneous local-global feature learning (dubbed 3DLG-Detector). 3DLG-Detector has two key contributions. First, it develops a Dynamic Points Interaction (DPI) module that preserves effective local features during pooling. Besides, DPI is detachable and can be incorporated into existing 3OD networks to boost their performance. Second, it develops a Global Context Aggregation module to aggregate multi-scale features from different layers of the encoder to achieve scene context-awareness. Our method shows improvements over thirteen competitors in terms of detection accuracy and robustness on both the SUN RGB-D and ScanNet datasets. Source code will be available upon publication.

Abstract— Geodesic paths and distances are among the most popular intrinsic properties of 3D surfaces. Traditionally, geodesic paths on discrete polygon surfaces were computed using shortest path algorithms, such as Dijkstra. However, such algorithms have two major limitations. They are non-differentiable which limits their direct usage in learnable pipelines and they are considerably time demanding. To address such limitations and alleviate the computational burden, we propose a learnable network to approximate geodesic paths. The proposed method is comprised by three major components: a graph neural network that encodes node positions in a high dimensional space, a path embedding that describes previously visited nodes and a point classifier that selects the next point in the path. The proposed method provides efficient approximations of the shortest paths and geodesic distances estimations. Given that all of the components of our method are fully differentiable, it can be directly plugged into any learnable pipeline as well as customized under any differentiable constraint. We extensively evaluate the proposed method with several qualitative and quantitative experiments.

Abstract— Neural Radiance Fields (NeRF) have emerged as a potent paradigm for representing scenes and synthesizing photo-realistic images. A main limitation of conventional NeRFs is that they often fail to produce high-quality renderings under novel viewpoints that are significantly different from the training viewpoints. In this paper, instead of exploiting few-shot image synthesis, we study the novel view extrapolation setting that (1) the training images can well describe an object, and (2) there is a notable discrepancy between the training and test viewpoints' distributions. We present RapNeRF (RAy Priors) as a solution. Our insight is that the inherent appearances of a 3D surface's arbitrary visible projections should be consistent. We thus propose a random ray casting policy that allows training unseen views using seen views. Furthermore, we show that a ray atlas pre-computed from the observed rays' viewing directions could further enhance the rendering quality for extrapolated views. A main limitation is that RapNeRF would remove the strong view-dependent effects because it leverages the multi-view consistency property.

Abstract— Dense correspondence between humans carries powerful semantic information that can be utilized to solve fundamental problems for full-body understanding such as in-the-wild surface matching, tracking and reconstruction. In this paper we present BodyMap, a new framework for obtaining high-definition full-body and continuous dense correspondence between in-the-wild images of clothed humans and the surface of a 3D template model. The correspondences cover fine details such as hands and hair, while capturing regions far from the body surface, such as loose clothing. Prior methods for estimating such dense surface correspondence i) cut a 3D body into parts which are unwrapped to a 2D UV space, producing discontinuities along part seams, or ii) use a single surface for representing the whole body, but none handled body details. Here, we introduce a novel network architecture with Vision Transformers that learn fine-level features on a continuous body surface. BodyMap outperforms prior work on various metrics and datasets, including DensePose-COCO by a large margin. Furthermore, we show various applications ranging from multi-layer dense cloth correspondence, neural rendering with novel-view synthesis and appearance swapping.

Abstract— Generative 3D face models are a powerful tool in computer vision. They provide pose and illumination invariance by modeling the space of 3D faces and the imaging process. The power of these models comes at the cost of an expensive and tedious construction process, which has led the community to focus on more easily constructed but less powerful models. With this paper we publish a generative 3D shape and texture model, the Basel face model (BFM), and demonstrate its application to several face recognition task. We improve on previous models by offering higher shape and texture accuracy due to a better scanning device and less correspondence artifacts due to an improved registration algorithm. The same 3D face model can be fit to 2D or 3D images acquired under different situations and with different sensors using an analysis by synthesis method. The resulting model parameters separate pose, lighting, imaging and identity parameters, which facilitates invariant face recognition across sensors and data sets by comparing only the identity parameters. We hope that the availability of this registered face model will spur research in generative models. Together with the model we publish a set of detailed recognition and reconstruction results on standard databases to allow complete algorithm comparisons.

Abstract— Automated 3D reconstruction of faces from images is challenging if the image material is difficult in terms of pose, lighting, occlusions and facial expressions, and if the initial 2D feature positions are inaccurate or unreliable. We propose a method that reconstructs individual 3D shapes from multiple single images of one person, judges their quality and then combines the best of all results. This is done separately for different regions of the face. The core element of this algorithm and the focus of our paper is a quality measure that judges a reconstruction without information about the true shape. We evaluate different quality measures, develop a method for combining results, and present a complete processing pipeline for automated reconstruction.

Abstract— We present a novel method for using Neural Networks (NNs) for finding solutions to a class of Partial Differential Equations (PDEs). Our method builds on recent advances in Neural Radiance Field research (NeRFs) and allows for a NN to converge to a PDE solution much faster than classic Physically Informed Neural Network (PINNs) approaches.

Abstract— In this paper, we present BEVerse, a unified framework for 3D perception and prediction based on multi-camera systems. Unlike existing studies focusing on the improvement of single-task approaches, BEVerse features in producing spatio-temporal Birds-Eye-View (BEV) representations from multi-camera videos and jointly reasoning about multiple tasks for vision-centric autonomous driving. Specifically, BEVerse first performs shared feature extraction and lifting to generate 4D BEV representations from multi-timestamp and multi-view images. After the ego-motion alignment, the spatio-temporal encoder is utilized for further feature extraction in BEV. Finally, multiple task decoders are attached for joint reasoning and prediction. Within the decoders, we propose the grid sampler to generate BEV features with different ranges and granularities for different tasks. Also, we design the method of iterative flow for memory-efficient future prediction. We show that the temporal information improves 3D object detection and semantic map construction, while the multi-task learning can implicitly benefit motion prediction. With extensive experiments on the nuScenes dataset, we show that the multi-task BEVerse outperforms existing single-task methods on 3D object detection, semantic map construction, and motion prediction. Compared with the sequential paradigm, BEVerse also favors in significantly improved efficiency. The code and trained models will be released at this https URL.

Abstract— Non-parametric face modeling aims to reconstruct 3D face only from images without shape assumptions. While plausible facial details are predicted, the models tend to over-depend on local color appearance and suffer from ambiguous noise. To address such problem, this paper presents a novel Learning to Aggregate and Personalize (LAP) framework for unsupervised robust 3D face modeling. Instead of using controlled environment, the proposed method implicitly disentangles ID-consistent and scene-specific face from unconstrained photo set. Specifically, to learn ID-consistent face, LAP adaptively aggregates intrinsic face factors of an identity based on a novel curriculum learning approach with relaxed consistency loss. To adapt the face for a personalized scene, we propose a novel attribute-refining network to modify ID-consistent face with target attribute and details. Based on the proposed method, we make unsupervised 3D face modeling benefit from meaningful image facial structure and possibly higher resolutions. Extensive experiments on benchmarks show LAP recovers superior or competitive face shape and texture, compared with state-of-the-art (SOTA) methods with or without prior and supervision.

Abstract— We are interested in learning data-driven representations that can generalize well, even when trained on inherently biased data. In particular, we face the case where some attributes (bias) of the data, if learned by the model, can severely compromise its generalization properties. We tackle this problem through the lens of information theory, leveraging recent findings for a differentiable estimation of mutual information. We propose a novel end-to-end optimization strategy, which simultaneously estimates and minimizes the mutual information between the learned representation and the data attributes. When applied on standard benchmarks, our model shows comparable or superior classification performance with respect to state-of-the-art approaches. Moreover, our method is general enough to be applicable to the problem of ``algorithmic fairness'', with competitive results.

Abstract— As a classic statistical model of 3D facial shape and texture, 3D Morphable Model (3DMM) is widely used in facial analysis, e.g., model fitting, image synthesis. Conventional 3DMM is learned from a set of well-controlled 2D face images with associated 3D face scans, and represented by two sets of PCA basis functions. Due to the type and amount of training data, as well as the linear bases, the representation power of 3DMM can be limited. To address these problems, this paper proposes an innovative framework to learn a nonlinear 3DMM model from a large set of unconstrained face images, without collecting 3D face scans. Specifically, given a face image as input, a network encoder estimates the projection, shape and texture parameters. Two decoders serve as the nonlinear 3DMM to map from the shape and texture parameters to the 3D shape and texture, respectively. With the projection parameter, 3D shape, and texture, a novel analytically-differentiable rendering layer is designed to reconstruct the original input face. The entire network is end-to-end trainable with only weak supervision. We demonstrate the superior representation power of our nonlinear 3DMM over its linear counterpart, and its contribution to face alignment and 3D reconstruction.

Abstract— Though Neural Radiance Field (NeRF) demonstrates compelling novel view synthesis results, it is still unintuitive to edit a pre-trained NeRF because the neural network's parameters and the scene geometry/appearance are often not explicitly associated. In this paper, we introduce the first framework that enables users to remove unwanted objects or retouch undesired regions in a 3D scene represented by a pre-trained NeRF without any category-specific data and training. The user first draws a free-form mask to specify a region containing unwanted objects over a rendered view from the pre-trained NeRF. Our framework first transfers the user-provided mask to other rendered views and estimates guiding color and depth images within these transferred masked regions. Next, we formulate an optimization problem that jointly inpaints the image content in all masked regions across multiple views by updating the NeRF model's parameters. We demonstrate our framework on diverse scenes and show it obtained visual plausible and structurally consistent results across multiple views using shorter time and less user manual efforts.

Abstract— We propose a Transformer-based NeRF (TransNeRF) to learn a generic neural radiance field conditioned on observed-view images for the novel view synthesis task. By contrast, existing MLP-based NeRFs are not able to directly receive observed views with an arbitrary number and require an auxiliary pooling-based operation to fuse source-view information, resulting in the missing of complicated relationships between source views and the target rendering view. Furthermore, current approaches process each 3D point individually and ignore the local consistency of a radiance field scene representation. These limitations potentially can reduce their performance in challenging real-world applications where large differences between source views and a novel rendering view may exist. To address these challenges, our TransNeRF utilizes the attention mechanism to naturally decode deep associations of an arbitrary number of source views into a coordinate-based scene representation. Local consistency of shape and appearance are considered in the ray-cast space and the surrounding-view space within a unified Transformer network. Experiments demonstrate that our TransNeRF, trained on a wide variety of scenes, can achieve better performance in comparison to state-of-the-art image-based neural rendering methods in both scene-agnostic and per-scene finetuning scenarios especially when there is a considerable gap between source views and a rendering view.

Abstract— Fully unsupervised 3D representation learning has gained attention owing to its advantages in data collection. A successful approach involves a viewpoint-aware approach that learns an image distribution based on generative models (e.g., generative adversarial networks (GANs)) while generating various view images based on 3D-aware models (e.g., neural radiance fields (NeRFs)). However, they require images with various views for training, and consequently, their application to datasets with few or limited viewpoints remains a challenge. As a complementary approach, an aperture rendering GAN (AR-GAN) that employs a defocus cue was proposed. However, an AR-GAN is a CNN-based model and represents a defocus independently from a viewpoint change despite its high correlation, which is one of the reasons for its performance. As an alternative to an AR-GAN, we propose an aperture rendering NeRF (AR-NeRF), which can utilize viewpoint and defocus cues in a unified manner by representing both factors in a common ray-tracing framework. Moreover, to learn defocus-aware and defocus-independent representations in a disentangled manner, we propose aperture randomized training, for which we learn to generate images while randomizing the aperture size and latent codes independently. During our experiments, we applied AR-NeRF to various natural image datasets, including flower, bird, and face images, the results of which demonstrate the utility of AR-NeRF for unsupervised learning of the depth and defocus effects.

Abstract— We present a method for the accurate 3D reconstruction of partly-symmetric objects. We build on the strengths of recent advances in neural reconstruction and rendering such as Neural Radiance Fields (NeRF). A major shortcoming of such approaches is that they fail to reconstruct any part of the object which is not clearly visible in the training image, which is often the case for in-the-wild images and videos. When evidence is lacking, structural priors such as symmetry can be used to complete the missing information. However, exploiting such priors in neural rendering is highly non-trivial: while geometry and non-reflective materials may be symmetric, shadows and reflections from the ambient scene are not symmetric in general. To address this, we apply a soft symmetry constraint to the 3D geometry and material properties, having factored appearance into lighting, albedo colour and reflectivity. We evaluate our method on the recently introduced CO3D dataset, focusing on the car category due to the challenge of reconstructing highly-reflective materials. We show that it can reconstruct unobserved regions with high fidelity and render high-quality novel view images.

Abstract— Volumetric neural rendering methods, such as neural radiance fields (NeRFs), have enabled photo-realistic novel view synthesis. However, in their standard form, NeRFs do not support the editing of objects, such as a human head, within a scene. In this work, we propose RigNeRF, a system that goes beyond just novel view synthesis and enables full control of head pose and facial expressions learned from a single portrait video. We model changes in head pose and facial expressions using a deformation field that is guided by a 3D morphable face model (3DMM). The 3DMM effectively acts as a prior for RigNeRF that learns to predict only residuals to the 3DMM deformations and allows us to render novel (rigid) poses and (non-rigid) expressions that were not present in the input sequence. Using only a smartphone-captured short video of a subject for training, we demonstrate the effectiveness of our method on free view synthesis of a portrait scene with explicit head pose and expression controls. The project page can be found here: this http URL.

Abstract— High-fidelity reconstruction of fluids from sparse multiview RGB videos remains a formidable challenge due to the complexity of the underlying physics as well as complex occlusion and lighting in captures. Existing solutions either assume knowledge of obstacles and lighting, or only focus on simple fluid scenes without obstacles or complex lighting, and thus are unsuitable for real-world scenes with unknown lighting or arbitrary obstacles. We present the first method to reconstruct dynamic fluid by leveraging the governing physics (ie, Navier -Stokes equations) in an end-to-end optimization from sparse videos without taking lighting conditions, geometry information, or boundary conditions as input. We provide a continuous spatio-temporal scene representation using neural networks as the ansatz of density and velocity solution functions for fluids as well as the radiance field for static objects. With a hybrid architecture that separates static and dynamic contents, fluid interactions with static obstacles are reconstructed for the first time without additional geometry input or human labeling. By augmenting time-varying neural radiance fields with physics-informed deep learning, our method benefits from the supervision of images and physical priors. To achieve robust optimization from sparse views, we introduced a layer-by-layer growing strategy to progressively increase the network capacity. Using progressively growing models with a new regularization term, we manage to disentangle density-color ambiguity in radiance fields without overfitting. A pretrained density-to-velocity fluid model is leveraged in addition as the data prior to avoid suboptimal velocity which underestimates vorticity but trivially fulfills physical equations. Our method exhibits high-quality results with relaxed constraints and strong flexibility on a representative set of synthetic and real flow captures.

Abstract— Recent works have shown that 3D-aware GANs trained on unstructured single image collections can generate multiview images of novel instances. The key underpinnings to achieve this are a 3D radiance field generator and a volume rendering process. However, existing methods either cannot generate high-resolution images (e.g., up to 256X256) due to the high computation cost of neural volume rendering, or rely on 2D CNNs for image-space upsampling which jeopardizes the 3D consistency across different views. This paper proposes a novel 3D-aware GAN that can generate high resolution images (up to 1024X1024) while keeping strict 3D consistency as in volume rendering. Our motivation is to achieve super-resolution directly in the 3D space to preserve 3D consistency. We avoid the otherwise prohibitively-expensive computation cost by applying 2D convolutions on a set of 2D radiance manifolds defined in the recent generative radiance manifold (GRAM) approach, and apply dedicated loss functions for effective GAN training at high resolution. Experiments on FFHQ and AFHQv2 datasets show that our method can produce high-quality 3D-consistent results that significantly outperform existing methods.

Abstract— State-of-the-art 3D-aware generative models rely on coordinate-based MLPs to parameterize 3D radiance fields. While demonstrating impressive results, querying an MLP for every sample along each ray leads to slow rendering. Therefore, existing approaches often render low-resolution feature maps and process them with an upsampling network to obtain the final image. Albeit efficient, neural rendering often entangles viewpoint and content such that changing the camera pose results in unwanted changes of geometry or appearance. Motivated by recent results in voxel-based novel view synthesis, we investigate the utility of sparse voxel grid representations for fast and 3D-consistent generative modeling in this paper. Our results demonstrate that monolithic MLPs can indeed be replaced by 3D convolutions when combining sparse voxel grids with progressive growing, free space pruning and appropriate regularization. To obtain a compact representation of the scene and allow for scaling to higher voxel resolutions, our model disentangles the foreground object (modeled in 3D) from the background (modeled in 2D). In contrast to existing approaches, our method requires only a single forward pass to generate a full 3D scene. It hence allows for efficient rendering from arbitrary viewpoints while yielding 3D consistent results with high visual fidelity.

Recently, Neural Radiance Fields (NeRF) is revolutionizing the task of novel view synthesis (NVS) for its superior performance. However, NeRF and its variants generally require a lengthy per-scene training procedure, where a multi-layer perceptron (MLP) is fitted to the captured images. To remedy the challenge, the voxel-grid representation has been proposed to significantly speed up the training. However, these existing methods can only deal with static scenes. How to develop an efficient and accurate dynamic view synthesis method remains an open problem. Extending the methods for static scenes to dynamic scenes is not straightforward as both the scene geometry and appearance change over time. In this paper, built on top of the recent advances in voxel-grid optimization, we propose a fast deformable radiance field method to handle dynamic scenes. Our method consists of two modules. The first module adopts a deformation grid to store 3D dynamic features, and a light-weight MLP for decoding the deformation that maps a 3D point in observation space to the canonical space using the interpolated features. The second module contains a density and a color grid to model the geometry and density of the scene. The occlusion is explicitly modeled to further improve the rendering quality. Experimental results show that our method achieves comparable performance to D-NeRF using only 20 minutes for training, which is more than 70x faster than D-NeRF, clearly demonstrating the efficiency of our proposed method.

Abstract— Neural approximations of scalar and vector fields, such as signed distance functions and radiance fields, have emerged as accurate, high-quality representations. State-of-the-art results are obtained by conditioning a neural approximation with a lookup from trainable feature grids that take on part of the learning task and allow for smaller, more efficient neural networks. Unfortunately, these feature grids usually come at the cost of significantly increased memory consumption compared to stand-alone neural network models. We present a dictionary method for compressing such feature grids, reducing their memory consumption by up to 100x and permitting a multiresolution representation which can be useful for out-of-core streaming. We formulate the dictionary optimization as a vector-quantized auto-decoder problem which lets us learn end-to-end discrete neural representations in a space where no direct supervision is available and with dynamic topology and structure. Our source code will be available at this https URL.

Abstract— Novel view synthesis (NVS) is a challenging task requiring systems to generate photorealistic images of scenes from new viewpoints, where both quality and speed are important for applications. Previous image-based rendering (IBR) methods are fast, but have poor quality when input views are sparse. Recent Neural Radiance Fields (NeRF) and generalizable variants give impressive results but are not real-time. In our paper, we propose a generalizable NVS method with sparse inputs, called FWD, which gives high-quality synthesis in real-time. With explicit depth and differentiable rendering, it achieves competitive results to the SOTA methods with 130-1000x speedup and better perceptual quality. If available, we can seamlessly integrate sensor depth during either training or inference to improve image quality while retaining real-time speed. With the growing prevalence of depths sensors, we hope that methods making use of depth will become increasingly useful.

Abstract— Constructing colorized point clouds from mobile laser scanning and images is a fundamental work in surveying and mapping. It is also an essential prerequisite for building digital twins for smart cities. However, existing public datasets are either in relatively small scales or lack accurate geometrical and color ground truth. This paper documents a multisensorial dataset named PolyU-BPCoMA which is distinctively positioned towards mobile colorized mapping. The dataset incorporates resources of 3D LiDAR, spherical imaging, GNSS and IMU on a backpack platform. Color checker boards are pasted in each surveyed area as targets and ground truth data are collected by an advanced terrestrial laser scanner (TLS). 3D geometrical and color information can be recovered in the colorized point clouds produced by the backpack system and the TLS, respectively. Accordingly, we provide an opportunity to benchmark the mapping and colorization accuracy simultaneously for a mobile multisensorial system. The dataset is approximately 800 GB in size covering both indoor and outdoor environments. The dataset and development kits are available at this https URL.

Abstract— Change detection (CD) aims to identify changes that occur in an image pair taken different times. Prior methods devise specific networks from scratch to predict change masks in pixel-level, and struggle with general segmentation problems. In this paper, we propose a new paradigm that reduces CD to semantic segmentation which means tailoring an existing and powerful semantic segmentation network to solve CD. This new paradigm conveniently enjoys the mainstream semantic segmentation techniques to deal with general segmentation problems in CD. Hence we can concentrate on studying how to detect changes. We propose a novel and importance insight that different change types exist in CD and they should be learned separately. Based on it, we devise a module named MTF to extract the change information and fuse temporal features. MTF enjoys high interpretability and reveals the essential characteristic of CD. And most segmentation networks can be adapted to solve the CD problems with our MTF module. Finally, we propose C-3PO, a network to detect changes at pixel-level. C-3PO achieves state-of-the-art performance without bells and whistles. It is simple but effective and can be considered as a new baseline in this field. Our code will be available.

Abstract— Object detection and semantic segmentation with the 3D lidar point cloud data require expensive annotation. We propose a data augmentation method that takes advantage of already annotated data multiple times. We propose an augmentation framework that reuses real data, automatically finds suitable placements in the scene to be augmented, and handles occlusions explicitly. Due to the usage of the real data, the scan points of newly inserted objects in augmentation sustain the physical characteristics of the lidar, such as intensity and raydrop. The pipeline proves competitive in training top-performing models for 3D object detection and semantic segmentation. The new augmentation provides a significant performance gain in rare and essential classes, notably 6.65% average precision gain for "Hard" pedestrian class in KITTI object detection or 2.14 mean IoU gain in the SemanticKITTI segmentation challenge over the state of the art.

Abstract— Recently, transformers have shown great potential in image classification and established state-of-the-art results on the ImageNet benchmark. However, compared to CNNs, transformers converge slowly and are prone to overfitting in low-data regimes due to the lack of spatial inductive biases. Such spatial inductive biases can be especially beneficial since the 2D structure of an input image is not well preserved in transformers. In this work, we present Spatial Prior-enhanced Self-Attention (SP-SA), a novel variant of vanilla Self-Attention (SA) tailored for vision transformers. Spatial Priors (SPs) are our proposed family of inductive biases that highlight certain groups of spatial relations. Unlike convolutional inductive biases, which are forced to focus exclusively on hard-coded local regions, our proposed SPs are learned by the model itself and take a variety of spatial relations into account. Specifically, the attention score is calculated with emphasis on certain kinds of spatial relations at each head, and such learned spatial foci can be complementary to each other. Based on SP-SA we propose the SP-ViT family, which consistently outperforms other ViT models with similar GFlops or parameters. Our largest model SP-ViT-L achieves a record-breaking 86.3% Top-1 accuracy with a reduction in the number of parameters by almost 50% compared to previous state-of-the-art model (150M for SP-ViT-L vs 271M for CaiT-M-36) among all ImageNet-1K models trained on 224x224 and fine-tuned on 384x384 resolution w/o extra data.

Abstract— Audio-visual automatic speech recognition (AV-ASR) is an extension of ASR that incorporates visual cues, often from the movements of a speaker's mouth. Unlike works that simply focus on the lip motion, we investigate the contribution of entire visual frames (visual actions, objects, background etc.). This is particularly useful for unconstrained videos, where the speaker is not necessarily visible. To solve this task, we propose a new sequence-to-sequence AudioVisual ASR TrAnsformeR (AVATAR) which is trained end-to-end from spectrograms and full-frame RGB. To prevent the audio stream from dominating training, we propose different word-masking strategies, thereby encouraging our model to pay attention to the visual stream. We demonstrate the contribution of the visual modality on the How2 AV-ASR benchmark, especially in the presence of simulated noise, and show that our model outperforms all other prior work by a large margin. Finally, we also create a new, real-world test bed for AV-ASR called VisSpeech, which demonstrates the contribution of the visual modality under challenging audio conditions.

Abstract— Lighter and faster models are crucial for the deployment of video super-resolution (VSR) on resource-limited devices, e.g., smartphones and wearable devices. In this paper, we develop Residual Sparsity Connection Learning (RSCL), a structured pruning scheme, to reduce the redundancy of convolution kernels and obtain a compact VSR network with a minor performance drop. However, residual blocks require the pruned filter indices of skip and residual connections to be the same, which is tricky for pruning. Thus, to mitigate the pruning restrictions of residual blocks, we design a Residual Sparsity Connection (RSC) scheme by preserving the feature channels and only operating on the important channels. Moreover, for the pixel-shuffle operation, we design a special pruning scheme by grouping several filters as pruning units to guarantee the accuracy of feature channel-space conversion after pruning. In addition, we introduce Temporal Finetuning (TF) to reduce the pruning error amplification of hidden states with temporal propagation. Extensive experiments show that the proposed RSCL significantly outperforms recent methods quantitatively and qualitatively. Codes and models will be released.

Abstract— Deep learning models have achieved remarkable success in different areas of machine learning over the past decade; however, the size and complexity of these models make them difficult to understand. In an effort to make them more interpretable, several recent works focus on explaining parts of a deep neural network through human-interpretable, semantic attributes. However, it may be impossible to completely explain complex models using only semantic attributes. In this work, we propose to augment these attributes with a small set of uninterpretable features. Specifically, we develop a novel explanation framework ELUDE (Explanation via Labelled and Unlabelled DEcomposition) that decomposes a model's prediction into two parts: one that is explainable through a linear combination of the semantic attributes, and another that is dependent on the set of uninterpretable features. By identifying the latter, we are able to analyze the "unexplained" portion of the model, obtaining insights into the information used by the model. We show that the set of unlabelled features can generalize to multiple models trained with the same feature space and compare our work to two popular attribute-oriented methods, Interpretable Basis Decomposition and Concept Bottleneck, and discuss the additional insights ELUDE provides.

Abstract— Predicting and anticipating future outcomes or reasoning about missing information in a sequence are critical skills for agents to be able to make intelligent decisions. This requires strong, temporally coherent generative capabilities. Diffusion models have shown remarkable success in several generative tasks, but have not been extensively explored in the video domain. We present Random-Mask Video Diffusion (RaMViD), which extends image diffusion models to videos using 3D convolutions, and introduces a new conditioning technique during training. By varying the mask we condition on, the model is able to perform video prediction, infilling, and upsampling. Due to our simple conditioning scheme, we can utilize the same architecture as used for unconditional training, which allows us to train the model in a conditional and unconditional fashion at the same time. We evaluate the model on two benchmark datasets for video prediction, on which we achieve state-of-the-art results, and one for video generation.

Abstract— The large computing and memory cost of deep neural networks (DNNs) often precludes their use in resource-constrained devices. Quantizing the parameters and operations to lower bit-precision offers substantial memory and energy savings for neural network inference, facilitating the use of DNNs on edge computing platforms. Recent efforts at quantizing DNNs have employed a range of techniques encompassing progressive quantization, step-size adaptation, and gradient scaling. This paper proposes a new quantization approach for mixed precision convolutional neural networks (CNNs) targeting edge-computing. Our method establishes a new pareto frontier in model accuracy and memory footprint demonstrating a range of quantized models, delivering best-in-class accuracy below 4.3 MB of weights (wgts.) and activations (acts.). Our main contributions are: (i) hardware-aware heterogeneous differentiable quantization with tensor-sliced learned precision, (ii) targeted gradient modification for wgts. and acts. to mitigate quantization errors, and (iii) a multi-phase learning schedule to address instability in learning arising from updates to the learned quantizer and model parameters. We demonstrate the effectiveness of our techniques on the ImageNet dataset across a range of models including EfficientNet-Lite0 (e.g., 4.14MB of wgts. and acts. at 67.66% accuracy) and MobileNetV2 (e.g., 3.51MB wgts. and acts. at 65.39% accuracy).

Abstract— Understanding to what extent neural networks memorize training data is an intriguing question with practical and theoretical implications. In this paper we show that in some cases a significant fraction of the training data can in fact be reconstructed from the parameters of a trained neural network classifier. We propose a novel reconstruction scheme that stems from recent theoretical results about the implicit bias in training neural networks with gradient-based methods. To the best of our knowledge, our results are the first to show that reconstructing a large portion of the actual training samples from a trained neural network classifier is generally possible. This has negative implications on privacy, as it can be used as an attack for revealing sensitive training data. We demonstrate our method for binary MLP classifiers on a few standard computer vision datasets.

Abstract— The emergence of neural networks has revolutionized the field of motion synthesis. Yet, learning to unconditionally synthesize motions from a given distribution remains a challenging task, especially when the motions are highly diverse. We present MoDi, an unconditional generative model that synthesizes diverse motions. Our model is trained in a completely unsupervised setting from a diverse, unstructured and unlabeled motion dataset and yields a well-behaved, highly semantic latent space. The design of our model follows the prolific architecture of StyleGAN and adapts two of its key technical components into the motion domain: a set of style-codes injected into each level of the generator hierarchy and a mapping function that learns and forms a disentangled latent space. We show that despite the lack of any structure in the dataset, the latent space can be semantically clustered, and facilitates semantic editing and motion interpolation. In addition, we propose a technique to invert unseen motions into the latent space, and demonstrate latent-based motion editing operations that otherwise cannot be achieved by naive manipulation of explicit motion representations. Our qualitative and quantitative experiments show that our framework achieves state-of-the-art synthesis quality that can follow the distribution of highly diverse motion datasets. Code and trained models will be released at this https URL.

Abstract— Longitudinal variations and complementary information inherent in longitudinal and multi-modal data play an important role in Alzheimer's disease (AD) prediction, particularly in identifying subjects with mild cognitive impairment who are about to have AD. However, longitudinal and multi-modal data may have missing data, which hinders the effective application of these data. Additionally, previous longitudinal studies require existing longitudinal data to achieve prediction, but AD prediction is expected to be conducted at patients' baseline visit (BL) in clinical practice. Thus, we proposed a multi-view imputation and cross-attention network (MCNet) to integrate data imputation and AD prediction in a unified framework and achieve accurate AD prediction. First, a multi-view imputation method combined with adversarial learning, which can handle a wide range of missing data situations and reduce imputation errors, was presented. Second, two cross-attention blocks were introduced to exploit the potential associations in longitudinal and multi-modal data. Finally, a multi-task learning model was built for data imputation, longitudinal classification, and AD prediction tasks. When the model was properly trained, the disease progression information learned from longitudinal data can be leveraged by BL data to improve AD prediction. The proposed method was tested on two independent testing sets and single-model data at BL to verify its effectiveness and flexibility on AD prediction. Results showed that MCNet outperformed several state-of-the-art methods. Moreover, the interpretability of MCNet was presented. Thus, our MCNet is a tool with a great application potential in longitudinal and multi-modal data analysis for AD prediction. Codes are available at this https URL.

Abstract— Despite the considerable progress in automatic abdominal multi-organ segmentation from CT/MRI scans in recent years, a comprehensive evaluation of the models' capabilities is hampered by the lack of a large-scale benchmark from diverse clinical scenarios. Constraint by the high cost of collecting and labeling 3D medical data, most of the deep learning models to date are driven by datasets with a limited number of organs of interest or samples, which still limits the power of modern deep models and makes it difficult to provide a fully comprehensive and fair estimate of various methods. To mitigate the limitations, we present AMOS, a large-scale, diverse, clinical dataset for abdominal organ segmentation. AMOS provides 500 CT and 100 MRI scans collected from multi-center, multi-vendor, multi-modality, multi-phase, multi-disease patients, each with voxel-level annotations of 15 abdominal organs, providing challenging examples and test-bed for studying robust segmentation algorithms under diverse targets and scenarios. We further benchmark several state-of-the-art medical segmentation models to evaluate the status of the existing methods on this new challenging dataset. We have made our datasets, benchmark servers, and baselines publicly available, and hope to inspire future research. Information can be found at this https URL.

Abstract— Locomotion in Virtual Reality (VR) is an important part of VR applications. Many scientists are enriching the community with different variations that enable locomotion in VR. Some of the most promising methods are gesture-based and do not require additional handheld hardware. Recent work focused mostly on user preference and performance of the different locomotion techniques. This ignores the learning effect that users go through while new methods are being explored. In this work, it is investigated whether and how quickly users can adapt to a hand gesture-based locomotion system in VR. Four different locomotion techniques are implemented and tested by participants. The goal of this paper is twofold: First, it aims to encourage researchers to consider the learning effect in their studies. Second, this study aims to provide insight into the learning effect of users in gesture-based systems.

Abstract— We propose a learning-based method to reconstruct the local terrain for locomotion with a mobile robot traversing urban environments. Using a stream of depth measurements from the onboard cameras and the robot's trajectory, the algorithm estimates the topography in the robot's vicinity. The raw measurements from these cameras are noisy and only provide partial and occluded observations that in many cases do not show the terrain the robot stands on. Therefore, we propose a 3D reconstruction model that faithfully reconstructs the scene, despite the noisy measurements and large amounts of missing data coming from the blind spots of the camera arrangement. The model consists of a 4D fully convolutional network on point clouds that learns the geometric priors to complete the scene from the context and an auto-regressive feedback to leverage spatio-temporal consistency and use evidence from the past. The network can be solely trained with synthetic data, and due to extensive augmentation, it is robust in the real world, as shown in the validation on a quadrupedal robot, ANYmal, traversing challenging settings. We run the pipeline on the robot's onboard low-power computer using an efficient sparse tensor implementation and show that the proposed method outperforms classical map representations.

Abstract— Alzheimer's disease (AD) is the most common cause of dementia. An early detection is crucial for slowing down the disease and mitigating risks related to the progression. While the combination of MRI and FDG-PET is the best image-based tool for diagnosis, FDG-PET is not always available. The reliable detection of Alzheimer's disease with only MRI could be beneficial, especially in regions where FDG-PET might not be affordable for all patients. To this end, we propose a multi-task method based on U-Net that takes T1-weighted MR images as an input to generate synthetic FDG-PET images and classifies the dementia progression of the patient into cognitive normal (CN), cognitive impairment (MCI), and AD. The attention gates used in both task heads can visualize the most relevant parts of the brain, guiding the examiner and adding interpretability. Results show the successful generation of synthetic FDG-PET images and a performance increase in disease classification over the naive single-task baseline.

Abstract— The detection of new multiple sclerosis (MS) lesions is an important marker of the evolution of the disease. The applicability of learning-based methods could automate this task efficiently. However, the lack of annotated longitudinal data with new-appearing lesions is a limiting factor for the training of robust and generalizing models. In this work, we describe a deep-learning-based pipeline addressing the challenging task of detecting and segmenting new MS lesions. First, we propose to use transfer-learning from a model trained on a segmentation task using single time-points. Therefore, we exploit knowledge from an easier task and for which more annotated datasets are available. Second, we propose a data synthesis strategy to generate realistic longitudinal time-points with new lesions using single time-point scans. In this way, we pretrain our detection model on large synthetic annotated datasets. Finally, we use a data-augmentation technique designed to simulate data diversity in MRI. By doing that, we increase the size of the available small annotated longitudinal datasets. Our ablation study showed that each contribution lead to an enhancement of the segmentation accuracy. Using the proposed pipeline, we obtained the best score for the segmentation and the detection of new MS lesions in the MSSEG2 MICCAI challenge.

Abstract— Video capsule endoscopy is a hot topic in computer vision and medicine. Deep learning can have a positive impact on the future of video capsule endoscopy technology. It can improve the anomaly detection rate, reduce physicians' time for screening, and aid in real-world clinical analysis. CADx classification system for video capsule endoscopy has shown a great promise for further improvement. For example, detection of cancerous polyp and bleeding can lead to swift medical response and improve the survival rate of the patients. To this end, an automated CADx system must have high throughput and decent accuracy. In this paper, we propose FocalConvNet, a focal modulation network integrated with lightweight convolutional layers for the classification of small bowel anatomical landmarks and luminal findings. FocalConvNet leverages focal modulation to attain global context and allows global-local spatial interactions throughout the forward pass. Moreover, the convolutional block with its intrinsic inductive/learning bias and capacity to extract hierarchical features allows our FocalConvNet to achieve favourable results with high throughput. We compare our FocalConvNet with other SOTA on Kvasir-Capsule, a large-scale VCE dataset with 44,228 frames with 13 classes of different anomalies. Our proposed method achieves the weighted F1-score, recall and MCC} of 0.6734, 0.6373 and 0.2974, respectively outperforming other SOTA methodologies. Furthermore, we report the highest throughput of 148.02 images/second rate to establish the potential of FocalConvNet in a real-time clinical environment. The code of the proposed FocalConvNet is available at this https URL.

Abstract— We introduce SoundSpaces 2.0, a platform for on-the-fly geometry-based audio rendering for 3D environments. Given a 3D mesh of a real-world environment, SoundSpaces can generate highly realistic acoustics for arbitrary sounds captured from arbitrary microphone locations. Together with existing 3D visual assets, it supports an array of audio-visual research tasks, such as audio-visual navigation, mapping, source localization and separation, and acoustic matching. Compared to existing resources, SoundSpaces 2.0 has the advantages of allowing continuous spatial sampling, generalization to novel environments, and configurable microphone and material properties. To our best knowledge, this is the first geometry-based acoustic simulation that offers high fidelity and realism while also being fast enough to use for embodied learning. We showcase the simulator's properties and benchmark its performance against real-world audio measurements. In addition, through two downstream tasks covering embodied navigation and far-field automatic speech recognition, highlighting sim2real performance for the latter. SoundSpaces 2.0 is publicly available to facilitate wider research for perceptual systems that can both see and hear.

Abstract— The visual world can be parsimoniously characterized in terms of distinct entities with sparse interactions. Discovering this compositional structure in dynamic visual scenes has proven challenging for end-to-end computer vision approaches unless explicit instance-level supervision is provided. Slot-based models leveraging motion cues have recently shown great promise in learning to represent, segment, and track objects without direct supervision, but they still fail to scale to complex real-world multi-object videos. In an effort to bridge this gap, we take inspiration from human development and hypothesize that information about scene geometry in the form of depth signals can facilitate object-centric learning. We introduce SAVi++, an object-centric video model which is trained to predict depth signals from a slot-based video representation. By further leveraging best practices for model scaling, we are able to train SAVi++ to segment complex dynamic scenes recorded with moving cameras, containing both static and moving objects of diverse appearance on naturalistic backgrounds, without the need for segmentation supervision. Finally, we demonstrate that by using sparse depth signals obtained from LiDAR, SAVi++ is able to learn emergent object segmentation and tracking from videos in the real-world Waymo Open dataset.

Abstract— The CLIP network measures the similarity between natural text and images; in this work, we investigate the entanglement of the representation of word images and natural images in its image encoder. First, we find that the image encoder has an ability to match word images with natural images of scenes described by those words. This is consistent with previous research that suggests that the meaning and the spelling of a word might be entangled deep within the network. On the other hand, we also find that CLIP has a strong ability to match nonsense words, suggesting that processing of letters is separated from processing of their meaning. To explicitly determine whether the spelling capability of CLIP is separable, we devise a procedure for identifying representation subspaces that selectively isolate or eliminate spelling capabilities. We benchmark our methods against a range of retrieval tasks, and we also test them by measuring the appearance of text in CLIP-guided generated images. We find that our methods are able to cleanly separate spelling capabilities of CLIP from the visual processing of natural images.

Abstract— This brief technical report describes our submission to the Action Spotting SoccerNet Challenge 2022. The challenge was part of the CVPR 2022 ActivityNet Workshop. Our submission was based on a recently proposed method which focuses on increasing temporal precision via a densely sampled set of detection anchors. Due to its emphasis on temporal precision, this approach had shown significant improvements in the tight average-mAP metric. Tight average-mAP was used as the evaluation criterion for the challenge, and is defined using small temporal evaluation tolerances, thus being more sensitive to small temporal errors. In order to further improve results, here we introduce small changes in the pre- and post-processing steps, and also combine different input feature types via late fusion. These changes brought improvements that helped us achieve the first place in the challenge and also led to a new state-of-the-art on SoccerNet's test set when using the dataset's standard experimental protocol. This report briefly reviews the action spotting method based on dense detection anchors, then focuses on the modifications introduced for the challenge. We also describe the experimental protocols and training procedures we used, and finally present our results.

Abstract— Neural rendering can be used to reconstruct implicit representations of shapes without 3D supervision. However, current neural surface reconstruction methods have difficulty learning high-frequency details of shapes, so that the reconstructed shapes are often oversmoothed. We propose a novel method to improve the quality of surface reconstruction in neural rendering. We follow recent work to model surfaces as signed distance fields. First, we offer a derivation to analyze the relationship between the signed distance function, the volume density, the transparency function, and the weighting function used in the volume rendering equation. Second, we observe that attempting to jointly encode high-frequency and low frequency components in a single signed distance function leads to unstable optimization. We propose to decompose the signed distance function in a base function and a displacement function together with a coarse-to-fine strategy to gradually increase the high-frequency details. Finally, we propose to use an adaptive strategy that enables the optimization to focus on improving certain regions near the surface where the signed distance fields have artifacts. Our qualitative and quantitative results show that our method can reconstruct high-frequency surface details and obtain better surface reconstruction quality than the current state of the art. Code will be released at this https URL.

Abstract— In this paper we propose a generative adversarial network (GAN) framework to enhance the perceptual quality of compressed videos. Our framework includes attention and adaptation to different quantization parameters (QPs) in a single model. The attention module exploits global receptive fields that can capture and align long-range correlations between consecutive frames, which can be beneficial for enhancing perceptual quality of videos. The frame to be enhanced is fed into the deep network together with its neighboring frames, and in the first stage features at different depths are extracted. Then extracted features are fed into attention blocks to explore global temporal correlations, followed by a series of upsampling and convolution layers. Finally, the resulting features are processed by the QP-conditional adaptation module which leverages the corresponding QP information. In this way, a single model can be used to enhance adaptively to various QPs without requiring multiple models specific for every QP value, while having similar performance. Experimental results demonstrate the superior performance of the proposed PeQuENet compared with the state-of-the-art compressed video quality enhancement algorithms.

Abstract— Attributed graph clustering is one of the most fundamental tasks among graph learning field, the goal of which is to group nodes with similar representations into the same cluster without human annotations. Recent studies based on graph contrastive learning method have achieved remarkable results when exploit graph-structured data. However, most existing methods 1) do not directly address the clustering task, since the representation learning and clustering process are separated; 2) depend too much on data augmentation, which greatly limits the capability of contrastive learning; 3) ignore the contrastive message for clustering tasks, which adversely degenerate the clustering results. In this paper, we propose a Neighborhood Contrast Framework for Attributed Graph Clustering, namely NCAGC, seeking for conquering the aforementioned limitations. Specifically, by leveraging the Neighborhood Contrast Module, the representation of neighbor nodes will be 'push closer' and become clustering-oriented with the neighborhood contrast loss. Moreover, a Contrastive Self-Expression Module is built by minimizing the node representation before and after the self-expression layer to constraint the learning of self-expression matrix. All the modules of NCAGC are optimized in a unified framework, so the learned node representation contains clustering-oriented messages. Extensive experimental results on four attributed graph datasets demonstrate the promising performance of NCAGC compared with 16 state-of-the-art clustering methods. The code is available at this https URL.

Abstract— Online few-shot learning describes a setting where models are trained and evaluated on a stream of data while learning emerging classes. While prior work in this setting has achieved very promising performance on instance classification when learning from data-streams composed of a single indoor environment, we propose to extend this setting to consider object classification on a series of several indoor environments, which is likely to occur in applications such as robotics. Importantly, our setting, which we refer to as online few-shot continual learning, injects the well-studied issue of catastrophic forgetting into the few-shot online learning paradigm. In this work, we benchmark several existing methods and adapted baselines within our setting, and show there exists a trade-off between catastrophic forgetting and online performance. Our findings motivate the need for future work in this setting, which can achieve better online performance without catastrophic forgetting.

Abstract— We propose a motion forecasting model called BANet, which means Boundary-Aware Network, and it is a variant of LaneGCN. We believe that it is not enough to use only the lane centerline as input to obtain the embedding features of the vector map nodes. The lane centerline can only provide the topology of the lanes, and other elements of the vector map also contain rich information. For example, the lane boundary can provide traffic rule constraint information such as whether it is possible to change lanes which is very important. Therefore, we achieved better performance by encoding more vector map elements in the motion forecasting model.We report our results on the 2022 Argoverse2 Motion Forecasting challenge and rank 1st on the test leaderboard.

Abstract— Adversarial training (AT) is a simple yet effective defense against adversarial attacks to image classification systems, which is based on augmenting the training set with attacks that maximize the loss. However, the effectiveness of AT as a defense for video classification has not been thoroughly studied. Our first contribution is to show that generating optimal attacks for video requires carefully tuning the attack parameters, especially the step size. Notably, we show that the optimal step size varies linearly with the attack budget. Our second contribution is to show that using a smaller (sub-optimal) attack budget at training time leads to a more robust performance at test time. Based on these findings, we propose three defenses against attacks with variable attack budgets. The first one, Adaptive AT, is a technique where the attack budget is drawn from a distribution that is adapted as training iterations proceed. The second, Curriculum AT, is a technique where the attack budget is increased as training iterations proceed. The third, Generative AT, further couples AT with a denoising generative adversarial network to boost robust performance. Experiments on the UCF101 dataset demonstrate that the proposed methods improve adversarial robustness against multiple attack types.

Abstract— Building 3D perception systems for autonomous vehicles that do not rely on LiDAR is a critical research problem because of the high expense of LiDAR systems compared to cameras and other sensors. Current methods use multi-view RGB data collected from cameras around the vehicle and neurally "lift" features from the perspective images to the 2D ground plane, yielding a "bird's eye view" (BEV) feature representation of the 3D space around the vehicle. Recent research focuses on the way the features are lifted from images to the BEV plane. We instead propose a simple baseline model, where the "lifting" step simply averages features from all projected image locations, and find that it outperforms the current state-of-the-art in BEV vehicle segmentation. Our ablations show that batch size, data augmentation, and input resolution play a large part in performance. Additionally, we reconsider the utility of radar input, which has previously been either ignored or found non-helpful by recent works. With a simple RGB-radar fusion module, we obtain a sizable boost in performance, approaching the accuracy of a LiDAR-enabled system.

Abstract— Image set-based visual classification methods have achieved remarkable performance, via characterising the image set in terms of a non-singular covariance matrix on a symmetric positive definite (SPD) manifold. To adapt to complicated visual scenarios better, several Riemannian networks (RiemNets) for SPD matrix nonlinear processing have recently been studied. However, it is pertinent to ask, whether greater accuracy gains can be achieved by simply increasing the depth of RiemNets. The answer appears to be negative, as deeper RiemNets tend to lose generalization ability. To explore a possible solution to this issue, we propose a new architecture for SPD matrix learning. Specifically, to enrich the deep representations, we adopt SPDNet [1] as the backbone, with a stacked Riemannian autoencoder (SRAE) built on the tail. The associated reconstruction error term can make the embedding functions of both SRAE and of each RAE an approximate identity mapping, which helps to prevent the degradation of statistical information. We then insert several residual-like blocks with shortcut connections to augment the representational capacity of SRAE, and to simplify the training of a deeper network. The experimental evidence demonstrates that our DreamNet can achieve improved accuracy with increased depth of the network.

Abstract— Multimodal sentiment analysis in videos is a key task in many real-world applications, which usually requires integrating multimodal streams including visual, verbal and acoustic behaviors. To improve the robustness of multimodal fusion, some of the existing methods let different modalities communicate with each other and modal the crossmodal interaction via transformers. However, these methods only use the single-scale representations during the interaction but forget to exploit multi-scale representations that contain different levels of semantic information. As a result, the representations learned by transformers could be biased especially for unaligned multimodal data. In this paper, we propose a multi-scale cooperative multimodal transformer (MCMulT) architecture for multimodal sentiment analysis. On the whole, the "multi-scale" mechanism is capable of exploiting the different levels of semantic information of each modality which are used for fine-grained crossmodal interactions. Meanwhile, each modality learns its feature hierarchies via integrating the crossmodal interactions from multiple level features of its source modality. In this way, each pair of modalities progressively builds feature hierarchies respectively in a cooperative manner. The empirical results illustrate that our MCMulT model not only outperforms existing approaches on unaligned multimodal sequences but also has strong performance on aligned multimodal sequences.

Abstract— Image captioning is the process of automatically generating a description of an image in natural language. Image captioning is one of the significant challenges in image understanding since it requires not only recognizing salient objects in the image but also their attributes and the way they interact. The system must then generate a syntactically and semantically correct caption that describes the image content in natural language. With the significant progress in deep learning models and their ability to effectively encode large sets of images and generate correct sentences, several neural-based captioning approaches have been proposed recently, each trying to achieve better accuracy and caption quality. This paper introduces an encoder-decoder-based image captioning system in which the encoder extracts spatial features from the image using ResNet-101. This stage is followed by a refining model, which uses an attention-on-attention mechanism to extract the visual features of the target image objects, then determine their interactions. The decoder consists of an attention-based recurrent module and a reflective attention module, which collaboratively apply attention to the visual and textual features to enhance the decoder's ability to model long-term sequential dependencies. Extensive experiments performed on Flickr30K, show the effectiveness of the proposed approach and the high quality of the generated captions.

Abstract— Recent self-supervised learning (SSL) methods have shown impressive results in learning visual representations from unlabeled images. This paper aims to improve their performance further by utilizing the architectural advantages of the underlying neural network, as the current state-of-the-art visual pretext tasks for SSL do not enjoy the benefit, i.e., they are architecture-agnostic. In particular, we focus on Vision Transformers (ViTs), which have gained much attention recently as a better architectural choice, often outperforming convolutional networks for various visual tasks. The unique characteristic of ViT is that it takes a sequence of disjoint patches from an image and processes patch-level representations internally. Inspired by this, we design a simple yet effective visual pretext task, coined SelfPatch, for learning better patch-level representations. To be specific, we enforce invariance against each patch and its neighbors, i.e., each patch treats similar neighboring patches as positive samples. Consequently, training ViTs with SelfPatch learns more semantically meaningful relations among patches (without using human-annotated labels), which can be beneficial, in particular, to downstream tasks of a dense prediction type. Despite its simplicity, we demonstrate that it can significantly improve the performance of existing SSL methods for various visual tasks, including object detection and semantic segmentation. Specifically, SelfPatch significantly improves the recent self-supervised ViT, DINO, by achieving +1.3 AP on COCO object detection, +1.2 AP on COCO instance segmentation, and +2.9 mIoU on ADE20K semantic segmentation.

Abstract— Noisy labels collected with limited annotation cost prevent medical image segmentation algorithms from learning precise semantic correlations. Previous segmentation arts of learning with noisy labels merely perform a pixel-wise manner to preserve semantics, such as pixel-wise label correction, but neglect the pair-wise manner. In fact, we observe that the pair-wise manner capturing affinity relations between pixels can greatly reduce the label noise rate. Motivated by this observation, we present a novel perspective for noisy mitigation by incorporating both pixel-wise and pair-wise manners, where supervisions are derived from noisy class and affinity labels, respectively. Unifying the pixel-wise and pair-wise manners, we propose a robust Joint Class-Affinity Segmentation (JCAS) framework to combat label noise issues in medical image segmentation. Considering the affinity in pair-wise manner incorporates contextual dependencies, a differentiated affinity reasoning (DAR) module is devised to rectify the pixel-wise segmentation prediction by reasoning about intra-class and inter-class affinity relations. To further enhance the noise resistance, a class-affinity loss correction (CALC) strategy is designed to correct supervision signals via the modeled noise label distributions in class and affinity labels. Meanwhile, CALC strategy interacts the pixel-wise and pair-wise manners through the theoretically derived consistency regularization. Extensive experiments under both synthetic and real-world noisy labels corroborate the efficacy of the proposed JCAS framework with a minimum gap towards the upper bound performance. The source code is available at this https URL.

Abstract— Conventional domain adaptation (DA) techniques aim to improve domain transferability by learning domain-invariant representations; while concurrently preserving the task-discriminability knowledge gathered from the labeled source data. However, the requirement of simultaneous access to labeled source and unlabeled target renders them unsuitable for the challenging source-free DA setting. The trivial solution of realizing an effective original to generic domain mapping improves transferability but degrades task discriminability. Upon analyzing the hurdles from both theoretical and empirical standpoints, we derive novel insights to show that a mixup between original and corresponding translated generic samples enhances the discriminability-transferability trade-off while duly respecting the privacy-oriented source-free setting. A simple but effective realization of the proposed insights on top of the existing source-free DA approaches yields state-of-the-art performance with faster convergence. Beyond single-source, we also outperform multi-source prior-arts across both classification and semantic segmentation benchmarks.

Abstract— To understand the real world using various types of data, Artificial Intelligence (AI) is the most used technique nowadays. While finding the pattern within the analyzed data represents the main task. This is performed by extracting representative features step, which is proceeded using the statistical algorithms or using some specific filters. However, the selection of useful features from large-scale data represented a crucial challenge. Now, with the development of convolution neural networks (CNNs), the feature extraction operation has become more automatic and easier. CNNs allow to work on large-scale size of data, as well as cover different scenarios for a specific task. For computer vision tasks, convolutional networks are used to extract features also for the other parts of a deep learning model. The selection of a suitable network for feature extraction or the other parts of a DL model is not random work. So, the implementation of such a model can be related to the target task as well as the computational complexity of it. Many networks have been proposed and become the famous networks used for any DL models in any AI task. These networks are exploited for feature extraction or at the beginning of any DL model which is named backbones. A backbone is a known network trained in many other tasks before and demonstrates its effectiveness. In this paper, an overview of the existing backbones, e.g. VGGs, ResNets, DenseNet, etc, is given with a detailed description. Also, a couple of computer vision tasks are discussed by providing a review of each task regarding the backbones used. In addition, a comparison in terms of performance is also provided, based on the backbone used for each task.

Abstract— Fiducial markers have been broadly used to identify objects or embed messages that can be detected by a camera. Primarily, existing detection methods assume that markers are printed on ideally planar surfaces. Markers often fail to be recognized due to various imaging artifacts of optical/perspective distortion and motion blur. To overcome these limitations, we propose a novel deformable fiducial marker system that consists of three main parts: First, a fiducial marker generator creates a set of free-form color patterns to encode significantly large-scale information in unique visual codes. Second, a differentiable image simulator creates a training dataset of photorealistic scene images with the deformed markers, being rendered during optimization in a differentiable manner. The rendered images include realistic shading with specular reflection, optical distortion, defocus and motion blur, color alteration, imaging noise, and shape deformation of markers. Lastly, a trained marker detector seeks the regions of interest and recognizes multiple marker patterns simultaneously via inverse deformation transformation. The deformable marker creator and detector networks are jointly optimized via the differentiable photorealistic renderer in an end-to-end manner, allowing us to robustly recognize a wide range of deformable markers with high accuracy. Our deformable marker system is capable of decoding 36-bit messages successfully at ~29 fps with severe shape deformation. Results validate that our system significantly outperforms the traditional and data-driven marker methods. Our learning-based marker system opens up new interesting applications of fiducial markers, including cost-effective motion capture of the human body, active 3D scanning using our fiducial markers' array as structured light patterns, and robust augmented reality rendering of virtual objects on dynamic surfaces.

Abstract— Unsupervised Domain Adaptation demonstrates great potential to mitigate domain shifts by transferring models from labeled source domains to unlabeled target domains. While Unsupervised Domain Adaptation has been applied to a wide variety of complex vision tasks, only few works focus on lane detection for autonomous driving. This can be attributed to the lack of publicly available datasets. To facilitate research in these directions, we propose CARLANE, a 3-way sim-to-real domain adaptation benchmark for 2D lane detection. CARLANE encompasses the single-target datasets MoLane and TuLane and the multi-target dataset MuLane. These datasets are built from three different domains, which cover diverse scenes and contain a total of 163K unique images, 118K of which are annotated. In addition we evaluate and report systematic baselines, including our own method, which builds upon Prototypical Cross-domain Self-supervised Learning. We find that false positive and false negative rates of the evaluated domain adaptation methods are high compared to those of fully supervised baselines. This affirms the need for benchmarks such as CARLANE to further strengthen research in Unsupervised Domain Adaptation for lane detection. CARLANE, all evaluated models and the corresponding implementations are publicly available at this https URL.

Abstract— In recent years, crowd counting has become an important issue in computer vision. In most methods, the density maps are generated by convolving with a Gaussian kernel from the ground-truth dot maps which are marked around the center of human heads. Due to the fixed geometric structures in CNNs and indistinct head-scale information, the head features are obtained incompletely. Deformable convolution is proposed to exploit the scale-adaptive capabilities for CNN features in the heads. By learning the coordinate offsets of the sampling points, it is tractable to improve the ability to adjust the receptive field. However, the heads are not uniformly covered by the sampling points in the deformable convolution, resulting in loss of head information. To handle the non-uniformed sampling, an improved Normed-Deformable Convolution (\textit{i.e.,}NDConv) implemented by Normed-Deformable loss (\textit{i.e.,}NDloss) is proposed in this paper. The offsets of the sampling points which are constrained by NDloss tend to be more even. Then, the features in the heads are obtained more completely, leading to better performance. Especially, the proposed NDConv is a light-weight module which shares similar computation burden with Deformable Convolution. In the extensive experiments, our method outperforms state-of-the-art methods on ShanghaiTech A, ShanghaiTech B, UCF_QNRF, and UCF_CC_50 dataset, achieving 61.4, 7.8, 91.2, and 167.2 MAE, respectively. The code is available at this https URL.

Abstract— Flood disasters cause enormous social and economic losses. However, both traditional physical models and learning-based flood forecasting models require massive historical flood data to train the model parameters. When come to some new site that does not have sufficient historical data, the model performance will drop dramatically due to overfitting. This technical report presents a Flood Domain Adaptation Network (FloodDAN), a baseline of applying Unsupervised Domain Adaptation (UDA) to the flood forecasting problem. Specifically, training of FloodDAN includes two stages: in the first stage, we train a rainfall encoder and a prediction head to learn general transferable hydrological knowledge on large-scale source domain data; in the second stage, we transfer the knowledge in the pretrained encoder into the rainfall encoder of target domain through adversarial domain alignment. During inference, we utilize the target domain rainfall encoder trained in the second stage and the prediction head trained in the first stage to get flood forecasting predictions. Experimental results on Tunxi and Changhua flood dataset show that FloodDAN can perform flood forecasting effectively with zero target domain supervision. The performance of the FloodDAN is on par with supervised models that uses 450-500 hours of supervision.

Abstract— Few-Shot Learning (FSL) requires vision models to quickly adapt to brand-new classification tasks with a shift in task distribution. Understanding the difficulties posed by this task distribution shift is central to FSL. In this paper, we show that a simple channel-wise feature transformation may be the key to unraveling this secret from a channel perspective. When facing novel few-shot tasks in the test-time datasets, this transformation can greatly improve the generalization ability of learned image representations, while being agnostic to the choice of training algorithms and datasets. Through an in-depth analysis of this transformation, we find that the difficulty of representation transfer in FSL stems from the severe channel bias problem of image representations: channels may have different importance in different tasks, while convolutional neural networks are likely to be insensitive, or respond incorrectly to such a shift. This points out a core problem of the generalization ability of modern vision systems and needs further attention in the future. Our code is available at this https URL.

Abstract— Current end-to-end autonomous driving methods either run a controller based on a planned trajectory or perform control prediction directly, which have spanned two separately studied lines of research. Seeing their potential mutual benefits to each other, this paper takes the initiative to explore the combination of these two well-developed worlds. Specifically, our integrated approach has two branches for trajectory planning and direct control, respectively. The trajectory branch predicts the future trajectory, while the control branch involves a novel multi-step prediction scheme such that the relationship between current actions and future states can be reasoned. The two branches are connected so that the control branch receives corresponding guidance from the trajectory branch at each time step. The outputs from two branches are then fused to achieve complementary advantages. Our results are evaluated in the closed-loop urban driving setting with challenging scenarios using the CARLA simulator. Even with a monocular camera input, the proposed approach ranks first on the official CARLA Leaderboard, outperforming other complex candidates with multiple sensors or fusion mechanisms by a large margin. The source code and data will be made publicly available at this https URL.

Abstract— Few-shot classification aims to learn a model that can generalize well to new tasks when only a few labeled samples are available. To make use of unlabeled data that are more abundantly available in real applications, Ren et al. \shortcite{ren2018meta} propose a semi-supervised few-shot classification method that assigns an appropriate label to each unlabeled sample by a manually defined metric. However, the manually defined metric fails to capture the intrinsic property in data. In this paper, we propose a \textbf{S}elf-\textbf{A}daptive \textbf{L}abel \textbf{A}ugmentation approach, called \textbf{SALA}, for semi-supervised few-shot classification. A major novelty of SALA is the task-adaptive metric, which can learn the metric adaptively for different tasks in an end-to-end fashion. Another appealing feature of SALA is a progressive neighbor selection strategy, which selects unlabeled data with high confidence progressively through the training phase. Experiments demonstrate that SALA outperforms several state-of-the-art methods for semi-supervised few-shot classification on benchmark datasets.

Abstract— Video question answering (VideoQA) is a complex task that requires diverse multi-modal data for training. Manual annotation of question and answers for videos, however, is tedious and prohibits scalability. To tackle this problem, recent methods consider zero-shot settings with no manual annotation of visual question-answer. In particular, a promising approach adapts frozen autoregressive language models pretrained on Web-scale text-only data to multi-modal inputs. In contrast, we here build on frozen bidirectional language models (BiLM) and show that such an approach provides a stronger and cheaper alternative for zero-shot VideoQA. In particular, (i) we combine visual inputs with the frozen BiLM using light trainable modules, (ii) we train such modules using Web-scraped multi-modal data, and finally (iii) we perform zero-shot VideoQA inference through masked language modeling, where the masked text is the answer to a given question. Our proposed approach, FrozenBiLM, outperforms the state of the art in zero-shot VideoQA by a significant margin on a variety of datasets, including LSMDC-FiB, iVQA, MSRVTT-QA, MSVD-QA, ActivityNet-QA, TGIF-FrameQA, How2QA and TVQA. It also demonstrates competitive performance in the few-shot and fully-supervised setting. Our code and models will be made publicly available at this https URL.

Abstract— In seismic interpretation, pixel-level labels of various rock structures can be time-consuming and expensive to obtain. As a result, there oftentimes exists a non-trivial quantity of unlabeled data that is left unused simply because traditional deep learning methods rely on access to fully labeled volumes. To rectify this problem, contrastive learning approaches have been proposed that use a self-supervised methodology in order to learn useful representations from unlabeled data. However, traditional contrastive learning approaches are based on assumptions from the domain of natural images that do not make use of seismic context. In order to incorporate this context within contrastive learning, we propose a novel positive pair selection strategy based on the position of slices within a seismic volume. We show that the learnt representations from our method out-perform a state of the art contrastive learning methodology in a semantic segmentation task.

Abstract— Unlike RGB cameras that use visible light bands (384∼769 THz) and Lidar that use infrared bands (361∼331 THz), Radars use relatively longer wavelength radio bands (77∼81 GHz), resulting in robust measurements in adverse weathers. Unfortunately, existing Radar datasets only contain a relatively small number of samples compared to the existing camera and Lidar datasets. This may hinder the development of sophisticated data-driven deep learning techniques for Radar-based perception. Moreover, most of the existing Radar datasets only provide 3D Radar tensor (3DRT) data that contain power measurements along the Doppler, range, and azimuth dimensions. As there is no elevation information, it is challenging to estimate the 3D bounding box of an object from 3DRT. In this work, we introduce KAIST-Radar (K-Radar), a novel large-scale object detection dataset and benchmark that contains 35K frames of 4D Radar tensor (4DRT) data with power measurements along the Doppler, range, azimuth, and elevation dimensions, together with carefully annotated 3D bounding box labels of objects on the roads. K-Radar includes challenging driving conditions such as adverse weathers (fog, rain, and snow) on various road structures (urban, suburban roads, alleyways, and highways). In addition to the 4DRT, we provide auxiliary measurements from carefully calibrated high-resolution Lidars, surround stereo cameras, and RTK-GPS. We also provide 4DRT-based object detection baseline neural networks (baseline NNs) and show that the height information is crucial for 3D object detection. And by comparing the baseline NN with a similarly-structured Lidar-based neural network, we demonstrate that 4D Radar is a more robust sensor for adverse weather conditions. All codes are available at this https URL.

Abstract— Crowd understanding has aroused the widespread interest in vision domain due to its important practical significance. Unfortunately, there is no effort to explore crowd understanding in multi-modal domain that bridges natural language and computer vision. Referring expression comprehension (REF) is such a representative multi-modal task. Current REF studies focus more on grounding the target object from multiple distinctive categories in general scenarios. It is difficult to applied to complex real-world crowd understanding. To fill this gap, we propose a new challenging dataset, called RefCrowd, which towards looking for the target person in crowd with referring expressions. It not only requires to sufficiently mine the natural language information, but also requires to carefully focus on subtle differences between the target and a crowd of persons with similar appearance, so as to realize the fine-grained mapping from language to vision. Furthermore, we propose a Fine-grained Multi-modal Attribute Contrastive Network (FMAC) to deal with REF in crowd understanding. It first decomposes the intricate visual and language features into attribute-aware multi-modal features, and then captures discriminative but robustness fine-grained attribute features to effectively distinguish these subtle differences between similar persons. The proposed method outperforms existing state-of-the-art (SoTA) methods on our RefCrowd dataset and existing REF datasets. In addition, we implement an end-to-end REF toolbox for the deeper research in multi-modal domain. Our dataset and code can be available at: this https URL.

Abstract— Equipped with a wide span of sensors, predominant autonomous driving solutions are becoming more modular-oriented for safe system design. Though these sensors have laid a solid foundation, most massive-production solutions up to date still fall into L2 phase. Among these, this http URL comes to our sight, claiming one $999 aftermarket device mounted with a single camera and board inside owns the ability to handle L2 scenarios. Together with open-sourced software of the entire system released by this http URL, the project is named Openpilot. Is it possible? If so, how is it made possible? With curiosity in mind, we deep-dive into Openpilot and conclude that its key to success is the end-to-end system design instead of a conventional modular framework. The model is briefed as Supercombo, and it can predict the ego vehicle's future trajectory and other road semantics on the fly from monocular input. Unfortunately, the training process and massive amount of data to make all these work are not publicly available. To achieve an intensive investigation, we try to reimplement the training details and test the pipeline on public benchmarks. The refactored network proposed in this work is referred to as OP-Deepdive. For a fair comparison of our version to the original Supercombo, we introduce a dual-model deployment scheme to test the driving performance in the real world. Experimental results on nuScenes, Comma2k19, CARLA, and in-house realistic scenarios verify that a low-cost device can indeed achieve most L2 functionalities and be on par with the original Supercombo model. In this report, we would like to share our latest findings, shed some light on the new perspective of end-to-end autonomous driving from an industrial product-level side, and potentially inspire the community to continue improving the performance. Our code, benchmarks are at this https URL.

Abstract— The NuCLS dataset contains over 220.000 annotations of cell nuclei in breast cancers. We show how to use these data to create a multi-rater model with the MIScnn Framework to automate the analysis of cell nuclei. For the model creation, we use the widespread U-Net approach embedded in a pipeline. This pipeline provides besides the high performance convolution neural network, several preprocessor techniques and a extended data exploration. The final model is tested in the evaluation phase using a wide variety of metrics with a subsequent visualization. Finally, the results are compared and interpreted with the results of the NuCLS study. As an outlook, indications are given which are important for the future development of models in the context of cell nuclei.

Abstract— Filter pruning method introduces structural sparsity by removing selected filters and is thus particularly effective for reducing complexity. Previous works empirically prune networks from the point of view that filter with smaller norm contributes less to the final results. However, such criteria has been proven sensitive to the distribution of filters, and the accuracy may hard to recover since the capacity gap is fixed once pruned. In this paper, we propose a novel filter pruning method called Asymptotic Soft Cluster Pruning (ASCP), to identify the redundancy of network based on the similarity of filters. Each filter from over-parameterized network is first distinguished by clustering, and then reconstructed to manually introduce redundancy into it. Several guidelines of clustering are proposed to better preserve feature extraction ability. After reconstruction, filters are allowed to be updated to eliminate the effect caused by mistakenly selected. Besides, various decaying strategies of the pruning rate are adopted to stabilize the pruning process and improve the final performance as well. By gradually generating more identical filters within each cluster, ASCP can remove them through channel addition operation with almost no accuracy drop. Extensive experiments on CIFAR-10 and ImageNet datasets show that our method can achieve competitive results compared with many state-of-the-art algorithms.

Abstract— Roof-mounted spinning LiDAR sensors are widely used by autonomous vehicles. However, most semantic datasets and algorithms used for LiDAR sequence segmentation operate on 360∘ frames, causing an acquisition latency incompatible with real-time applications. To address this issue, we first introduce HelixNet, a 10 billion point dataset with fine-grained labels, timestamps, and sensor rotation information necessary to accurately assess the real-time readiness of segmentation algorithms. Second, we propose Helix4D, a compact and efficient spatio-temporal transformer architecture specifically designed for rotating LiDAR sequences. Helix4D operates on acquisition slices corresponding to a fraction of a full sensor rotation, significantly reducing the total latency. Helix4D reaches accuracy on par with the best segmentation algorithms on HelixNet and SemanticKITTI with a reduction of over 5× in terms of latency and 50× in model size. The code and data are available at: this https URL.

Abstract— Pyramidal networks are standard methods for multi-scale object detection. Current researches on feature pyramid networks usually adopt layer connections to collect features from certain levels of the feature hierarchy, and do not consider the significant differences among them. We propose a better architecture of feature pyramid networks, named selective multi-scale learning (SMSL), to address this issue. SMSL is efficient and general, which can be integrated in both single-stage and two-stage detectors to boost detection performance, with nearly no extra inference cost. RetinaNet combined with SMSL obtains 1.8% improvement in AP (from 39.1% to 40.9%) on COCO dataset. When integrated with SMSL, two-stage detectors can get around 1.0% improvement in AP.

Abstract— In this paper, we introduce an enormous dataset HaGRID (HAnd Gesture Recognition Image Dataset) for hand gesture recognition (HGR) systems. This dataset contains 552,992 samples divided into 18 classes of gestures. The annotations consist of bounding boxes of hands with gesture labels and markups of leading hands. The proposed dataset allows for building HGR systems, which can be used in video conferencing services, home automation systems, the automotive sector, services for people with speech and hearing impairments, etc. We are especially focused on interaction with devices to manage them. That is why all 18 chosen gestures are functional, familiar to the majority of people, and may be an incentive to take some action. In addition, we used crowdsourcing platforms to collect the dataset and took into account various parameters to ensure data diversity. We describe the challenges of using existing HGR datasets for our task and provide a detailed overview of them. Furthermore, the baselines for the hand detection and gesture classification tasks are proposed.

Abstract— Visual domain adaptation (DA) seeks to transfer trained models to unseen, unlabeled domains across distribution shift, but approaches typically focus on adapting convolutional neural network architectures initialized with supervised ImageNet representations. In this work, we shift focus to adapting modern architectures for object recognition -- the increasingly popular Vision Transformer (ViT) -- and modern pretraining based on self-supervised learning (SSL). Inspired by the design of recent SSL approaches based on learning from partial image inputs generated via masking or cropping -- either by learning to predict the missing pixels, or learning representational invariances to such augmentations -- we propose PACMAC, a simple two-stage adaptation algorithm for self-supervised ViTs. PACMAC first performs in-domain SSL on pooled source and target data to learn task-discriminative features, and then probes the model's predictive consistency across a set of partial target inputs generated via a novel attention-conditioned masking strategy, to identify reliable candidates for self-training. Our simple approach leads to consistent performance gains over competing methods that use ViTs and self-supervised initializations on standard object recognition benchmarks. Code available at this https URL.

Abstract— Online knowledge distillation conducts knowledge transfer among all student models to alleviate the reliance on pre-trained models. However, existing online methods rely heavily on the prediction distributions and neglect the further exploration of the representational knowledge. In this paper, we propose a novel Multi-scale Feature Extraction and Fusion method (MFEF) for online knowledge distillation, which comprises three key components: Multi-scale Feature Extraction, Dual-attention and Feature Fusion, towards generating more informative feature maps for distillation. The multiscale feature extraction exploiting divide-and-concatenate in channel dimension is proposed to improve the multi-scale representation ability of feature maps. To obtain more accurate information, we design a dual-attention to strengthen the important channel and spatial regions adaptively. Moreover, we aggregate and fuse the former processed feature maps via feature fusion to assist the training of student models. Extensive experiments on CIF AR-10, CIF AR-100, and CINIC-10 show that MFEF transfers more beneficial representational knowledge for distillation and outperforms alternative methods among various network architectures.

Abstract— Object detection has made substantial progress in the last decade, due to the capability of convolution in extracting local context of objects. However, the scales of objects are diverse and current convolution can only process single-scale input. The capability of traditional convolution with a fixed receptive field in dealing with such a scale variance problem, is thus limited. Multi-scale feature representation has been proven to be an effective way to mitigate the scale variance problem. Recent researches mainly adopt partial connection with certain scales, or aggregate features from all scales and focus on the global information across the scales. However, the information across spatial and depth dimensions is ignored. Inspired by this, we propose the multi-scale convolution (MSConv) to handle this problem. Taking into consideration scale, spatial and depth information at the same time, MSConv is able to process multi-scale input more comprehensively. MSConv is effective and computationally efficient, with only a small increase of computational cost. For most of the single-stage object detectors, replacing the traditional convolutions with MSConvs in the detection head can bring more than 2.5% improvement in AP (on COCO 2017 dataset), with only 3% increase of FLOPs. MSConv is also flexible and effective for two-stage object detectors. When extended to the mainstream two-stage object detectors, MSConv can bring up to 3.0% improvement in AP. Our best model under single-scale testing achieves 48.9% AP on COCO 2017 test-dev split, which surpasses many state-of-the-art methods.

Abstract— Neural networks for image classification tasks assume that any given image during inference belongs to one of the training classes. This closed-set assumption is challenged in real-world applications where models may encounter inputs of unknown classes. Open-set recognition aims to solve this problem by rejecting unknown classes while classifying known classes correctly. In this paper, we propose to utilize gradient-based representations obtained from a known classifier to train an unknown detector with instances of known classes only. Gradients correspond to the amount of model updates required to properly represent a given sample, which we exploit to understand the model's capability to characterize inputs with its learned features. Our approach can be utilized with any classifier trained in a supervised manner on known classes without the need to model the distribution of unknown samples explicitly. We show that our gradient-based approach outperforms state-of-the-art methods by up to 11.6% in open-set classification.

Abstract— Though the state-of-the architectures for semantic segmentation, such as HRNet, demonstrate impressive accuracy, the complexity arising from their salient design choices hinders a range of model acceleration tools, and further they make use of operations that are inefficient on current hardware. This paper demonstrates that a simple encoder-decoder architecture with a ResNet-like backbone and a small multi-scale head, performs on-par or better than complex semantic segmentation architectures such as HRNet, FANet and DDRNets. Naively applying deep backbones designed for Image Classification to the task of Semantic Segmentation leads to sub-par results, owing to a much smaller effective receptive field of these backbones. Implicit among the various design choices put forth in works like HRNet, DDRNet, and FANet are networks with a large effective receptive field. It is natural to ask if a simple encoder-decoder architecture would compare favorably if comprised of backbones that have a larger effective receptive field, though without the use of inefficient operations like dilated convolutions. We show that with minor and inexpensive modifications to ResNets, enlarging the receptive field, very simple and competitive baselines can be created for Semantic Segmentation. We present a family of such simple architectures for desktop as well as mobile targets, which match or exceed the performance of complex models on the Cityscapes dataset. We hope that our work provides simple yet effective baselines for practitioners to develop efficient semantic segmentation models.

Abstract— The aetiology of head and neck squamous cell carcinoma (HNSCC) involves multiple carcinogens such as alcohol, tobacco and infection with human papillomavirus (HPV). As the HPV infection influences the prognosis, treatment and survival of patients with HNSCC, it is important to determine the HPV status of these tumours. In this paper, we propose a novel triplet-ranking loss function and a multiple instance learning pipeline for HPV status prediction. This achieves a new state-of-the-art performance in HPV detection using only the routine H&E stained WSIs on two HNSCC cohorts. Furthermore, a comprehensive tumour microenvironment profiling was performed, which characterised the unique patterns between HPV+/- HNSCC from genomic, immunology and cellular perspectives. Positive correlations of the proposed score with different subtypes of T cells (e.g. T cells follicular helper, CD8+ T cells), and negative correlations with macrophages and connective cells (e.g. fibroblast) were identified, which is in line with clinical findings. Unique gene expression profiles were also identified with respect to HPV infection status, and is in line with existing findings.

Abstract— Adversarial attacks in deep learning models, especially for safety-critical systems, are gaining more and more attention in recent years, due to the lack of trust in the security and robustness of AI models. Yet the more primitive adversarial attacks might be physically infeasible or require some resources that are hard to access like the training data, which motivated the emergence of patch attacks. In this survey, we provide a comprehensive overview to cover existing techniques of adversarial patch attacks, aiming to help interested researchers quickly catch up with the progress in this field. We also discuss existing techniques for developing detection and defences against adversarial patches, aiming to help the community better understand this field and its applications in the real world.

Abstract— Learning visual representations through self-supervision is an extremely challenging task as the network needs to sieve relevant patterns from spurious distractors without the active guidance provided by supervision. This is achieved through heavy data augmentation, large-scale datasets and prohibitive amounts of compute. Video self-supervised learning (SSL) suffers from added challenges: video datasets are typically not as large as image datasets, compute is an order of magnitude larger, and the amount of spurious patterns the optimizer has to sieve through is multiplied several fold. Thus, directly learning self-supervised representations from video data might result in sub-optimal performance. To address this, we propose to utilize a strong image-based model, pre-trained with self- or language supervision, in a video representation learning framework, enabling the model to learn strong spatial and temporal information without relying on the video labeled data. To this end, we modify the typical video-based SSL design and objective to encourage the video encoder to \textit{subsume} the semantic content of an image-based model trained on a general domain. The proposed algorithm is shown to learn much more efficiently (i.e. in less epochs and with a smaller batch) and results in a new state-of-the-art performance on standard downstream tasks among single-modality SSL methods.

Abstract— We present a system for realistic one-shot mesh-based human head avatars creation, ROME for short. Using a single photograph, our model estimates a person-specific head mesh and the associated neural texture, which encodes both local photometric and geometric details. The resulting avatars are rigged and can be rendered using a neural network, which is trained alongside the mesh and texture estimators on a dataset of in-the-wild videos. In the experiments, we observe that our system performs competitively both in terms of head geometry recovery and the quality of renders, especially for the cross-person reenactment. See results this https URL.

Abstract— Image super-resolution is an important research area in computer vision that has a wide variety of applications including surveillance, medical imaging etc. Real-world signal image super-resolution has become very popular now-a-days due to its real-time application. There are still a lot of scopes to improve real-world single image super-resolution specially during challenging weather scenarios. In this paper, we have proposed a new algorithm to perform real-world single image super-resolution during rainy condition. Our proposed method can mitigate the influence of rainy conditions during image super-resolution. Our experiment results show that our proposed algorithm can perform image super-resolution decreasing the negative effects of the rain.

Abstract— By leveraging contrastive learning, clustering, and other pretext tasks, unsupervised methods for learning image representations have reached impressive results on standard benchmarks. The result has been a crowded field - many methods with substantially different implementations yield results that seem nearly identical on popular benchmarks, such as linear evaluation on ImageNet. However, a single result does not tell the whole story. In this paper, we compare methods using performance-based benchmarks such as linear evaluation, nearest neighbor classification, and clustering for several different datasets, demonstrating the lack of a clear front-runner within the current state-of-the-art. In contrast to prior work that performs only supervised vs. unsupervised comparison, we compare several different unsupervised methods against each other. To enrich this comparison, we analyze embeddings with measurements such as uniformity, tolerance, and centered kernel alignment (CKA), and propose two new metrics of our own: nearest neighbor graph similarity and linear prediction overlap. We reveal through our analysis that in isolation, single popular methods should not be treated as though they represent the field as a whole, and that future work ought to consider how to leverage the complimentary nature of these methods. We also leverage CKA to provide a framework to robustly quantify augmentation invariance, and provide a reminder that certain types of invariance will be undesirable for downstream tasks.

Abstract— Transformer-based architectures have become competitive across a variety of visual domains, most notably images and videos. While prior work has studied these modalities in isolation, having a common architecture suggests that one can train a single unified model for multiple visual modalities. Prior attempts at unified modeling typically use architectures tailored for vision tasks, or obtain worse performance compared to single modality models. In this work, we show that masked autoencoding can be used to train a simple Vision Transformer on images and videos, without requiring any labeled data. This single model learns visual representations that are comparable to or better than single-modality representations on both image and video benchmarks, while using a much simpler architecture. In particular, our single pretrained model can be finetuned to achieve 86.5% on ImageNet and 75.3% on the challenging Something Something-v2 video benchmark. Furthermore, this model can be learned by dropping 90% of the image and 95% of the video patches, enabling extremely fast training.

Abstract— Existing GAN inversion and editing methods work well for aligned objects with a clean background, such as portraits and animal faces, but often struggle for more difficult categories with complex scene layouts and object occlusions, such as cars, animals, and outdoor images. We propose a new method to invert and edit such complex images in the latent space of GANs, such as StyleGAN2. Our key idea is to explore inversion with a collection of layers, spatially adapting the inversion process to the difficulty of the image. We learn to predict the "invertibility" of different image segments and project each segment into a latent layer. Easier regions can be inverted into an earlier layer in the generator's latent space, while more challenging regions can be inverted into a later feature space. Experiments show that our method obtains better inversion results compared to the recent approaches on complex categories, while maintaining downstream editability. Please refer to our project page at this https URL.

Abstract— Data augmentation is a necessity to enhance data efficiency in deep learning. For vision-language pre-training, data is only augmented either for images or for text in previous works. In this paper, we present MixGen: a joint data augmentation for vision-language representation learning to further improve data efficiency. It generates new image-text pairs with semantic relationships preserved by interpolating images and concatenating text. It's simple, and can be plug-and-played into existing pipelines. We evaluate MixGen on four architectures, including CLIP, ViLT, ALBEF and TCL, across five downstream vision-language tasks to show its versatility and effectiveness. For example, adding MixGen in ALBEF pre-training leads to absolute performance improvements on downstream tasks: image-text retrieval (+6.2% on COCO fine-tuned and +5.3% on Flicker30K zero-shot), visual grounding (+0.9% on RefCOCO+), visual reasoning (+0.9% on NLVR$^{2}$), visual question answering (+0.3% on VQA2.0), and visual entailment (+0.4% on SNLI-VE).

Abstract— Capitalizing on the recent advances in image generation models, existing controllable face image synthesis methods are able to generate high-fidelity images with some levels of controllability, e.g., controlling the shapes, expressions, textures, and poses of the generated face images. However, these methods focus on 2D image generative models, which are prone to producing inconsistent face images under large expression and pose changes. In this paper, we propose a new NeRF-based conditional 3D face synthesis framework, which enables 3D controllability over the generated face images by imposing explicit 3D conditions from 3D face priors. At its core is a conditional Generative Occupancy Field (cGOF) that effectively enforces the shape of the generated face to commit to a given 3D Morphable Model (3DMM) mesh. To achieve accurate control over fine-grained 3D face shapes of the synthesized image, we additionally incorporate a 3D landmark loss as well as a volume warping loss into our synthesis algorithm. Experiments validate the effectiveness of the proposed method, which is able to generate high-fidelity face images and shows more precise 3D controllability than state-of-the-art 2D-based controllable face synthesis methods. Find code and demo at this https URL.

Abstract— We introduce a unified framework for group equivariant networks on homogeneous spaces derived from a Fourier perspective. We consider tensor-valued feature fields, before and after a convolutional layer. We present a unified derivation of kernels via the Fourier domain by leveraging the sparsity of Fourier coefficients of the lifted feature fields. The sparsity emerges when the stabilizer subgroup of the homogeneous space is a compact Lie group. We further introduce a nonlinear activation, via an elementwise nonlinearity on the regular representation after lifting and projecting back to the field through an equivariant convolution. We show that other methods treating features as the Fourier coefficients in the stabilizer subgroup are special cases of our activation. Experiments on SO(3) and SE(3) show state-of-the-art performance in spherical vector field regression, point cloud classification, and molecular completion.

Abstract— Recovering the spatial layout of the cameras and the geometry of the scene from extreme-view images is a longstanding challenge in computer vision. Prevailing 3D reconstruction algorithms often adopt the image matching paradigm and presume that a portion of the scene is co-visible across images, yielding poor performance when there is little overlap among inputs. In contrast, humans can associate visible parts in one image to the corresponding invisible components in another image via prior knowledge of the shapes. Inspired by this fact, we present a novel concept called virtual correspondences (VCs). VCs are a pair of pixels from two images whose camera rays intersect in 3D. Similar to classic correspondences, VCs conform with epipolar geometry; unlike classic correspondences, VCs do not need to be co-visible across views. Therefore VCs can be established and exploited even if images do not overlap. We introduce a method to find virtual correspondences based on humans in the scene. We showcase how VCs can be seamlessly integrated with classic bundle adjustment to recover camera poses across extreme views. Experiments show that our method significantly outperforms state-of-the-art camera pose estimation methods in challenging scenarios and is comparable in the traditional densely captured setup. Our approach also unleashes the potential of multiple downstream tasks such as scene reconstruction from multi-view stereo and novel view synthesis in extreme-view scenarios.

Abstract— Adapting to a continuously evolving environment is a safety-critical challenge inevitably faced by all autonomous driving systems. Existing image and video driving datasets, however, fall short of capturing the mutable nature of the real world. In this paper, we introduce the largest multi-task synthetic dataset for autonomous driving, SHIFT. It presents discrete and continuous shifts in cloudiness, rain and fog intensity, time of day, and vehicle and pedestrian density. Featuring a comprehensive sensor suite and annotations for several mainstream perception tasks, SHIFT allows investigating the degradation of a perception system performance at increasing levels of domain shift, fostering the development of continuous adaptation strategies to mitigate this problem and assess model robustness and generality. Our dataset and benchmark toolkit are publicly available at www.vis.xyz/shift.

Abstract— Capturing general deforming scenes is crucial for many computer graphics and vision applications, and it is especially challenging when only a monocular RGB video of the scene is available. Competing methods assume dense point tracks, 3D templates, large-scale training datasets, or only capture small-scale deformations. In contrast to those, our method, Ub4D, makes none of these assumptions while outperforming the previous state of the art in challenging scenarios. Our technique includes two new, in the context of non-rigid 3D reconstruction, components, i.e., 1) A coordinate-based and implicit neural representation for non-rigid scenes, which enables an unbiased reconstruction of dynamic scenes, and 2) A novel dynamic scene flow loss, which enables the reconstruction of larger deformations. Results on our new dataset, which will be made publicly available, demonstrate the clear improvement over the state of the art in terms of surface reconstruction accuracy and robustness to large deformations. Visit the project page this https URL.

Abstract— Blind face restoration is a highly ill-posed problem that often requires auxiliary guidance to 1) improve the mapping from degraded inputs to desired outputs, or 2) complement high-quality details lost in the inputs. In this paper, we demonstrate that a learned discrete codebook prior in a small proxy space largely reduces the uncertainty and ambiguity of restoration mapping by casting blind face restoration as a code prediction task, while providing rich visual atoms for generating high-quality faces. Under this paradigm, we propose a Transformer-based prediction network, named CodeFormer, to model global composition and context of the low-quality faces for code prediction, enabling the discovery of natural faces that closely approximate the target faces even when the inputs are severely degraded. To enhance the adaptiveness for different degradation, we also propose a controllable feature transformation module that allows a flexible trade-off between fidelity and quality. Thanks to the expressive codebook prior and global modeling, CodeFormer outperforms state of the arts in both quality and fidelity, showing superior robustness to degradation. Extensive experimental results on synthetic and real-world datasets verify the effectiveness of our method.

Abstract— Immersive displays such as VR headsets, AR glasses, Multiview displays, Free point televisions have emerged as a new class of display technologies in recent years, offering a better visual experience and viewer engagement as compared to conventional displays. With the evolution of 3D video and display technologies, the consumer market for High Dynamic Range (HDR) cameras and displays is quickly growing. The lack of appropriate experimental data is a critical hindrance for the development of primary research efforts in the field of 3D HDR video technology. Also, the unavailability of sufficient real world multi-exposure experimental dataset is a major bottleneck for HDR imaging research, thereby limiting the quality of experience (QoE) for the viewers. In this paper, we introduce a diversified stereoscopic multi-exposure dataset captured within the campus of Indian Institute of Technology Madras, which is home to a diverse flora and fauna. The dataset is captured using ZED stereoscopic camera and provides intricate scenes of outdoor locations such as gardens, roadside views, festival venues, buildings and indoor locations such as academic and residential areas. The proposed dataset accommodates wide depth range, complex depth structure, complicate object movement, illumination variations, rich color dynamics, texture discrepancy in addition to significant randomness introduced by moving camera and background motion. The proposed dataset is made publicly available to the research community. Furthermore, the procedure for capturing, aligning and calibrating multi-exposure stereo videos and images is described in detail. Finally, we have discussed the progress, challenges, potential use cases and future research opportunities with respect to HDR imaging, depth estimation, consistent tone mapping and 3D HDR coding.

Abstract— Visual question answering (VQA) in surgery is largely unexplored. Expert surgeons are scarce and are often overloaded with clinical and academic workloads. This overload often limits their time answering questionnaires from patients, medical students or junior residents related to surgical procedures. At times, students and junior residents also refrain from asking too many questions during classes to reduce disruption. While computer-aided simulators and recording of past surgical procedures have been made available for them to observe and improve their skills, they still hugely rely on medical experts to answer their questions. Having a Surgical-VQA system as a reliable 'second opinion' could act as a backup and ease the load on the medical experts in answering these questions. The lack of annotated medical data and the presence of domain-specific terms has limited the exploration of VQA for surgical procedures. In this work, we design a Surgical-VQA task that answers questionnaires on surgical procedures based on the surgical scene. Extending the MICCAI endoscopic vision challenge 2018 dataset and workflow recognition dataset further, we introduce two Surgical-VQA datasets with classification and sentence-based answers. To perform Surgical-VQA, we employ vision-text transformers models. We further introduce a residual MLP-based VisualBert encoder model that enforces interaction between visual and text tokens, improving performance in classification-based answering. Furthermore, we study the influence of the number of input image patches and temporal visual features on the model performance in both classification and sentence-based answering.

Abstract— Black-box attacks usually face two problems: poor transferability and the inability to evade the adversarial defense. To overcome these shortcomings, we create an original approach to generate adversarial examples by smoothing the linear structure of the texture in the benign image, called AdvSmo. We construct the adversarial examples without relying on any internal information to the target model and design the imperceptible-high attack success rate constraint to guide the Gabor filter to select appropriate angles and scales to smooth the linear texture from the input images to generate adversarial examples. Benefiting from the above design concept, AdvSmo will generate adversarial examples with strong transferability and solid evasiveness. Finally, compared to the four advanced black-box adversarial attack methods, for the eight target models, the results show that AdvSmo improves the average attack success rate by 9% on the CIFAR-10 and 16% on the Tiny-ImageNet dataset compared to the best of these attack methods.

Abstract— In this paper, we propose an end-to-end SpA-Former to recover a shadow-free image from a single shaded image. Unlike traditional methods that require two steps for shadow detection and then shadow removal, the SpA-Former unifies these steps into one, which is a one-stage network capable of directly learning the mapping function between shadows and no shadows, it does not require a separate shadow detection. Thus, SpA-former is adaptable to real image de-shadowing for shadows projected on different semantic regions. SpA-Former consists of transformer layer and a series of joint Fourier transform residual blocks and two-wheel joint spatial attention. The network in this paper is able to handle the task while achieving a very fast processing efficiency. Our code is relased on this https URL

Abstract— We present a novel pyramidal output representation to ensure parsimony with our "specialize and fuse" process for semantic segmentation. A pyramidal "output" representation consists of coarse-to-fine levels, where each level is "specialize" in a different class distribution (e.g., more stuff than things classes at coarser levels). Two types of pyramidal outputs (i.e., unity and semantic pyramid) are "fused" into the final semantic output, where the unity pyramid indicates unity-cells (i.e., all pixels in such cell share the same semantic label). The process ensures parsimony by predicting a relatively small number of labels for unity-cells (e.g., a large cell of grass) to build the final semantic output. In addition to the "output" representation, we design a coarse-to-fine contextual module to aggregate the "features" representation from different levels. We validate the effectiveness of each key module in our method through comprehensive ablation studies. Finally, our approach achieves state-of-the-art performance on three widely-used semantic segmentation datasets -- ADE20K, COCO-Stuff, and Pascal-Context.

Abstract— In this paper, we provide a detailed survey of 3D Morphable Face Models over the 20 years since they were first proposed. The challenges in building and applying these models, namely capture, modeling, image formation, and image analysis, are still active research topics, and we review the state-of-the-art in each of these areas. We also look ahead, identifying unsolved challenges, proposing directions for future research and highlighting the broad range of current and future applications.

Abstract— We study open-world 3D scene understanding, a family of tasks that require agents to reason about their 3D environment with an open-set vocabulary and out-of-domain visual inputs - a critical skill for robots to operate in the unstructured 3D world. Towards this end, we propose Semantic Abstraction (SemAbs), a framework that equips 2D Vision-Language Models (VLMs) with new 3D spatial capabilities, while maintaining their zero-shot robustness. We achieve this abstraction using relevancy maps extracted from CLIP, and learn 3D spatial and geometric reasoning skills on top of those abstractions in a semantic-agnostic manner. We demonstrate the usefulness of SemAbs on two open-world 3D scene understanding tasks: 1) completing partially observed objects and 2) localizing hidden objects from language descriptions. Experiments show that SemAbs can generalize to novel vocabulary, materials/lighting, classes, and domains (i.e., real-world scans) from training on limited 3D synthetic data. Code and data will be available at this https URL.

Abstract— The aim of change detection (CD) is to detect changes occurred in the same area by comparing two images of that place taken at different times. The challenging part of the CD is to keep track of the changes the user wants to highlight, such as new buildings, and to ignore changes due to external factors such as environmental, lighting condition, fog or seasonal changes. Recent developments in the field of deep learning enabled researchers to achieve outstanding performance in this area. In particular, different mechanisms of space-time attention allowed to exploit the spatial features that are extracted from the models and to correlate them also in a temporal way by exploiting both the available images. The downside is that the models have become increasingly complex and large, often unfeasible for edge applications. These are limitations when the models must be applied to the industrial field or in applications requiring real-time performances. In this work we propose a novel model, called TinyCD, demonstrating to be both lightweight and effective, able to achieve performances comparable or even superior to the current state of the art with 13-150X fewer parameters. In our approach we have exploited the importance of low-level features to compare images. To do this, we use only few backbone blocks. This strategy allow us to keep the number of network parameters low. To compose the features extracted from the two images, we introduce a novel, economical in terms of parameters, mixing block capable of cross correlating features in both space and time domains. Finally, to fully exploit the information contained in the computed features, we define the PW-MLP block able to perform a pixel wise classification. Source code, models and results are available here: this https URL.

Abstract— In this paper, we present an approach for minimizing the computational complexity of trained Convolutional Neural Networks (ConvNet). The idea is to approximate all elements of a given ConvNet and replace the original convolutional filters and parameters (pooling and bias coefficients; and activation function) with efficient approximations capable of extreme reductions in computational complexity. Low-complexity convolution filters are obtained through a binary (zero-one) linear programming scheme based on the Frobenius norm over sets of dyadic rationals. The resulting matrices allow for multiplication-free computations requiring only addition and bit-shifting operations. Such low-complexity structures pave the way for low-power, efficient hardware designs. We applied our approach on three use cases of different complexity: (i) a "light" but efficient ConvNet for face detection (with around 1000 parameters); (ii) another one for hand-written digit classification (with more than 180000 parameters); and (iii) a significantly larger ConvNet: AlexNet with ≈1.2 million matrices. We evaluated the overall performance on the respective tasks for different levels of approximations. In all considered applications, very low-complexity approximations have been derived maintaining an almost equal classification performance.

Abstract— Trajectory prediction using deep neural networks (DNNs) is an essential component of autonomous driving (AD) systems. However, these methods are vulnerable to adversarial attacks, leading to serious consequences such as collisions. In this work, we identify two key ingredients to defend trajectory prediction models against adversarial attacks including (1) designing effective adversarial training methods and (2) adding domain-specific data augmentation to mitigate the performance degradation on clean data. We demonstrate that our method is able to improve the performance by 46% on adversarial data and at the cost of only 3% performance degradation on clean data, compared to the model trained with clean data. Additionally, compared to existing robust methods, our method can improve performance by 21% on adversarial examples and 9% on clean data. Our robust model is evaluated with a planner to study its downstream impacts. We demonstrate that our model can significantly reduce the severe accident rates (e.g., collisions and off-road driving).

Abstract— The current COVID-19 pandemic is a serious threat to humanity that directly affects the lungs. Automatic identification of COVID-19 is a challenge for health care officials. The standard gold method for diagnosing COVID-19 is Reverse Transcription Polymerase Chain Reaction (RT-PCR) to collect swabs from affected people. Some limitations encountered while collecting swabs are related to accuracy and longtime duration. Chest CT (Computed Tomography) is another test method that helps healthcare providers quickly identify the infected lung areas. It was used as a supporting tool for identifying COVID-19 in an earlier stage. With the help of deep learning, the CT imaging characteristics of COVID-19. Researchers have proven it to be highly effective for COVID-19 CT image classification. In this study, we review the recent deep learning techniques that can use to detect the COVID-19 disease. Relevant studies were collected by various databases such as Web of Science, Google Scholar, and PubMed. Finally, we compare the results of different deep learning models, and CT image analysis is discussed.

Abstract— In this paper, we address the Online Unsupervised Domain Adaptation (OUDA) problem and propose a novel multi-stage framework to solve real-world situations when the target data are unlabeled and arriving online sequentially in batches. To project the data from the source and the target domains to a common subspace and manipulate the projected data in real-time, our proposed framework institutes a novel method, called an Incremental Computation of Mean-Subspace (ICMS) technique, which computes an approximation of mean-target subspace on a Grassmann manifold and is proven to be a close approximate to the Karcher mean. Furthermore, the transformation matrix computed from the mean-target subspace is applied to the next target data in the recursive-feedback stage, aligning the target data closer to the source domain. The computation of transformation matrix and the prediction of next-target subspace leverage the performance of the recursive-feedback stage by considering the cumulative temporal dependency among the flow of the target subspace on the Grassmann manifold. The labels of the transformed target data are predicted by the pre-trained source classifier, then the classifier is updated by the transformed data and predicted labels. Extensive experiments on six datasets were conducted to investigate in depth the effect and contribution of each stage in our proposed framework and its performance over previous approaches in terms of classification accuracy and computational speed. In addition, the experiments on traditional manifold-based learning models and neural-network-based learning models demonstrated the applicability of our proposed framework for various types of learning models.

Abstract— Recovering the 3D motion of the heart from cine cardiac magnetic resonance (CMR) imaging enables the assessment of regional myocardial function and is important for understanding and analyzing cardiovascular disease. However, 3D cardiac motion estimation is challenging because the acquired cine CMR images are usually 2D slices which limit the accurate estimation of through-plane motion. To address this problem, we propose a novel multi-view motion estimation network (MulViMotion), which integrates 2D cine CMR images acquired in short-axis and long-axis planes to learn a consistent 3D motion field of the heart. In the proposed method, a hybrid 2D/3D network is built to generate dense 3D motion fields by learning fused representations from multi-view images. To ensure that the motion estimation is consistent in 3D, a shape regularization module is introduced during training, where shape information from multi-view images is exploited to provide weak supervision to 3D motion estimation. We extensively evaluate the proposed method on 2D cine CMR images from 580 subjects of the UK Biobank study for 3D motion tracking of the left ventricular myocardium. Experimental results show that the proposed method quantitatively and qualitatively outperforms competing methods.

Abstract— In this study, a method has been developed to improve the resolution of histological human placenta images. For this purpose, a paired series of high- and low-resolution images have been collected to train a deep neural network model that can predict image residuals required to improve the resolution of the input images. A modified version of the U-net neural network model has been tailored to find the relationship between the low resolution and residual images. After training for 900 epochs on an augmented dataset of 1000 images, the relative mean squared error of 0.003 is achieved for the prediction of 320 test images. The proposed method has not only improved the contrast of the low-resolution images at the edges of cells but added critical details and textures that mimic high-resolution images of placenta villous space.

Abstract— In this study, we have tailored a pixel tracking method for temporal extrapolation of the ventricular segmentation masks in cardiac magnetic resonance images. The pixel tracking process starts from the end-diastolic frame of the heart cycle using the available manually segmented images to predict the end-systolic segmentation mask. The superpixels approach is used to divide the raw images into smaller cells and in each time frame, new labels are assigned to the image cells which leads to tracking the movement of the heart wall elements through different frames. The tracked masks at the end of systole are compared with the already available manually segmented masks and dice scores are found to be between 0.81 to 0.84. Considering the fact that the proposed method does not necessarily require a training dataset, it could be an attractive alternative approach to deep learning segmentation methods in scenarios where training data are limited.

Abstract— In the practical applications of computed tomography imaging, the projection data may be acquired within a limited-angle range and corrupted by noises due to the limitation of scanning conditions. The noisy incomplete projection data results in the ill-posedness of the inverse problems. In this work, we theoretically verify that the low-resolution reconstruction problem has better numerical stability than the high-resolution problem. In what follows, a novel low-resolution image prior based CT reconstruction model is proposed to make use of the low-resolution image to improve the reconstruction quality. More specifically, we build up a low-resolution reconstruction problem on the down-sampled projection data, and use the reconstructed low-resolution image as prior knowledge for the original limited-angle CT problem. We solve the constrained minimization problem by the alternating direction method with all subproblems approximated by the convolutional neural networks. Numerical experiments demonstrate that our double-resolution network outperforms both the variational method and popular learning-based reconstruction methods on noisy limited-angle reconstruction problems.

Abstract— The MultiEarth 2022 Image-to-Image Translation challenge provides a well-constrained test bed for generating the corresponding RGB Sentinel-2 imagery with the given Sentinel-1 VV & VH imagery. In this challenge, we designed various generation models and found the SPADE [1] and pix2pixHD [2] models could perform our best results. In our self-evaluation, the SPADE-2 model with L1-loss can achieve 0.02194 MAE score and 31.092 PSNR dB. In our final submission, the best model can achieve 0.02795 MAE score ranked No.1 on the leader board.

Abstract— In ultrasound imaging the appearance of homogeneous regions of tissue is subject to speckle, which for certain applications can make the detection of tissue irregularities difficult. To cope with this, it is common practice to apply speckle reduction filters to the images. Most conventional filtering techniques are fairly hand-crafted and often need to be finely tuned to the present hardware, imaging scheme and application. Learning based techniques on the other hand suffer from the need for a target image for training (in case of fully supervised techniques) or require narrow, complex physics-based models of the speckle appearance that might not apply in all cases. With this work we propose a deep-learning based method for speckle removal without these limitations. To enable this, we make use of realistic ultrasound simulation techniques that allow for instantiation of several independent speckle realizations that represent the exact same tissue, thus allowing for the application of image reconstruction techniques that work with pairs of differently corrupted data. Compared to two other state-of-the-art approaches (non-local means and the Optimized Bayesian non-local means filter) our method performs favorably in qualitative comparisons and quantitative evaluation, despite being trained on simulations alone, and is several orders of magnitude faster.

Abstract— In this paper, we introduce DA$^{2}$, the first large-scale dual-arm dexterity-aware dataset for the generation of optimal bimanual grasping pairs for arbitrary large objects. The dataset contains about 9M pairs of parallel-jaw grasps, generated from more than 6000 objects and each labeled with various grasp dexterity measures. In addition, we propose an end-to-end dual-arm grasp evaluation model trained on the rendered scenes from this dataset. We utilize the evaluation model as our baseline to show the value of this novel and nontrivial dataset by both online analysis and real robot experiments. All data and related code will be open-sourced at this https URL.

Abstract— Generalizability of deep learning models may be severely affected by the difference in the distributions of the train (source domain) and the test (target domain) sets, e.g., when the sets are produced by different hardware. As a consequence of this domain shift, a certain model might perform well on data from one clinic, and then fail when deployed in another. We propose a very light and transparent approach to perform test-time domain adaptation. The idea is to substitute the target low-frequency Fourier space components that are deemed to reflect the style of an image. To maximize the performance, we implement the "optimal style donor" selection technique, and use a number of source data points for altering a single target scan appearance (Multi-Source Transferring). We study the effect of severity of domain shift on the performance of the method, and show that our training-free approach reaches the state-of-the-art level of complicated deep domain adaptation models. The code for our experiments is released.

Abstract— A gaze-fixating robot perceives distance to the fixated object and relative positions of surrounding objects immediately, accurately, and robustly. We show how fixation, which is the act of looking at one object while moving, exploits regularities in the geometry of 3D space to obtain this information. These regularities introduce rotation-translation couplings that are not commonly used in structure from motion. To validate, we use a Franka Emika Robot with an RGB camera. We a) find that error in distance estimate is less than 5 mm at a distance of 15 cm, and b) show how relative position can be used to find obstacles under challenging scenarios. We combine accurate distance estimates and obstacle information into a reactive robot behavior that is able to pick up objects of unknown size, while impeded by unforeseen obstacles.

Abstract— Prostate cancer is the most dangerous cancer diagnosed in men worldwide. Prostate diagnosis has been affected by many factors, such as lesion complexity, observer visibility, and variability. Many techniques based on Magnetic Resonance Imaging (MRI) have been used for prostate cancer identification and classification in the last few decades. Developing these techniques is crucial and has a great medical effect because they improve the treatment benefits and the chance of patients' survival. A new technique that depends on MRI has been proposed to improve the diagnosis. This technique consists of two stages. First, the MRI images have been preprocessed to make the medical image more suitable for the detection step. Second, prostate cancer identification has been performed based on a pre-trained deep learning model, InceptionResNetV2, that has many advantages and achieves effective results. In this paper, the InceptionResNetV2 deep learning model used for this purpose has average accuracy equals to 89.20%, and the area under the curve (AUC) equals to 93.6%. The experimental results of this proposed new deep learning technique represent promising and effective results compared to other previous techniques.

Abstract— Detecting out-of-distribution (OOD) data at inference time is crucial for many applications of machine learning. We present XOOD: a novel extreme value-based OOD detection framework for image classification that consists of two algorithms. The first, XOOD-M, is completely unsupervised, while the second XOOD-L is self-supervised. Both algorithms rely on the signals captured by the extreme values of the data in the activation layers of the neural network in order to distinguish between in-distribution and OOD instances. We show experimentally that both XOOD-M and XOOD-L outperform state-of-the-art OOD detection methods on many benchmark data sets in both efficiency and accuracy, reducing false-positive rate (FPR95) by 50%, while improving the inferencing time by an order of magnitude.

Abstract— The accurate and consistent border segmentation plays an important role in the tumor volume estimation and its treatment in the field of Medical Image Segmentation. Globally, Lung cancer is one of the leading causes of death and the early detection of lung nodules is essential for the early cancer diagnosis and survival rate of patients. The goal of this study was to demonstrate the feasibility of Deephealth toolkit including PyECVL and PyEDDL libraries to precisely segment lung nodules. Experiments for lung nodules segmentation has been carried out on UniToChest using PyECVL and PyEDDL, for data pre-processing as well as neural network training. The results depict accurate segmentation of lung nodules across a wide diameter range and better accuracy over a traditional detection approach. The datasets and the code used in this paper are publicly available as a baseline reference.

Abstract— The CT perfusion (CTP) is a medical exam for measuring the passage of a bolus of contrast solution through the brain on a pixel-by-pixel basis. The objective is to draw "perfusion maps" (namely cerebral blood volume, cerebral blood flow and time to peak) very rapidly for ischemic lesions, and to be able to distinguish between core and penumubra regions. A precise and quick diagnosis, in a context of ischemic stroke, can determine the fate of the brain tissues and guide the intervention and treatment in emergency conditions. In this work we present UniToBrain dataset, the very first open-source dataset for CTP. It comprises a cohort of more than a hundred of patients, and it is accompanied by patients metadata and ground truth maps obtained with state-of-the-art algorithms. We also propose a novel neural networks-based algorithm, using the European library ECVL and EDDL for the image processing and developing deep learning models respectively. The results obtained by the neural network models match the ground truth and open the road towards potential sub-sampling of the required number of CT maps, which impose heavy radiation doses to the patients.

Abstract— Convolutional neural networks (CNNs) have been the de facto standard in a diverse set of computer vision tasks for many years. Especially, deep neural networks based on seminal architectures such as U-shaped models with skip-connections or atrous convolution with pyramid pooling have been tailored to a wide range of medical image analysis tasks. The main advantage of such architectures is that they are prone to detaining versatile local features. However, as a general consensus, CNNs fail to capture long-range dependencies and spatial correlations due to the intrinsic property of confined receptive field size of convolution operations. Alternatively, Transformer, profiting from global information modelling that stems from the self-attention mechanism, has recently attained remarkable performance in natural language processing and computer vision. Nevertheless, previous studies prove that both local and global features are critical for a deep model in dense prediction, such as segmenting complicated structures with disparate shapes and configurations. To this end, this paper proposes TransDeepLab, a novel DeepLab-like pure Transformer for medical image segmentation. Specifically, we exploit hierarchical Swin-Transformer with shifted windows to extend the DeepLabv3 and model the Atrous Spatial Pyramid Pooling (ASPP) module. A thorough search of the relevant literature yielded that we are the first to model the seminal DeepLab model with a pure Transformer-based model. Extensive experiments on various medical image segmentation tasks verify that our approach performs superior or on par with most contemporary works on an amalgamation of Vision Transformer and CNN-based methods, along with a significant reduction of model complexity. The codes and trained models are publicly available at this https URL.

Abstract— Synthesizing multi-character interactions is a challenging task due to the complex and varied interactions between the characters. In particular, precise spatiotemporal alignment between characters is required in generating close interactions such as dancing and fighting. Existing work in generating multi-character interactions focuses on generating a single type of reactive motion for a given sequence which results in a lack of variety of the resultant motions. In this paper, we propose a novel way to create realistic human reactive motions which are not presented in the given dataset by mixing and matching different types of close interactions. We propose a Conditional Hierarchical Generative Adversarial Network with Multi-Hot Class Embedding to generate the Mix and Match reactive motions of the follower from a given motion sequence of the leader. Experiments are conducted on both noisy (depth-based) and high-quality (MoCap-based) interaction datasets. The quantitative and qualitative results show that our approach outperforms the state-of-the-art methods on the given datasets. We also provide an augmented dataset with realistic reactive motions to stimulate future research in this area. The code is available at this https URL.

Abstract— Current deep learning adaptive optimizer methods adjust the step magnitude of parameter updates by altering the effective learning rate used by each parameter. Motivated by the known inverse relation between batch size and learning rate on update step magnitudes, we introduce a novel training procedure that dynamically decides the dimension and the composition of the current update step. Our procedure, Dynamic Batch Adaptation (DBA) analyzes the gradients of every sample and selects the subset that best improves certain metrics such as gradient variance for each layer of the network. We present results showing DBA significantly improves the speed of model convergence. Additionally, we find that DBA produces an increased improvement over standard optimizers when used in data scarce conditions where, in addition to convergence speed, it also significantly improves model generalization, managing to train a network with a single fully connected hidden layer using only 1% of the MNIST dataset to reach 97.79% test accuracy. In an even more extreme scenario, it manages to reach 97.44% test accuracy using only 10 samples per class. These results represent a relative error rate reduction of 81.78% and 88.07% respectively, compared to the standard optimizers, Stochastic Gradient Descent (SGD) and Adam.

Abstract— A powerful category of data poisoning attacks modify a subset of training examples by small adversarial perturbations to change the prediction of certain test-time data. Existing defense mechanisms are not desirable to deploy in practice, as they often drastically harm the generalization performance, or are attack-specific and prohibitively slow to apply. Here, we propose a simple but highly effective approach that unlike existing methods breaks various types of poisoning attacks with the slightest drop in the generalization performance. We make the key observation that attacks exploit sharp loss regions to craft adversarial perturbations which can substantially alter examples' gradient or representations under small perturbations. To break poisoning attacks, our approach comprises two components: an optimized friendly noise that is generated to maximally perturb examples without degrading the performance, and a random varying noise component. The first component takes examples farther away from the sharp loss regions, and the second component smooths out the loss landscape. The combination of both components builds a very light-weight but extremely effective defense against the most powerful triggerless targeted and hidden-trigger backdoor poisoning attacks, including Gradient Matching, Bulls-eye Polytope, and Sleeper Agent. We show that our friendly noise is transferable to other architectures, and adaptive attacks cannot break our defense due to its random noise component.

Abstract— The recent increase in popularity of volumetric representations for scene reconstruction and novel view synthesis has put renewed focus on animating volumetric content at high visual quality and in real-time. While implicit deformation methods based on learned functions can produce impressive results, they are `black boxes' to artists and content creators, they require large amounts of training data to generalise meaningfully, and they do not produce realistic extrapolations outside the training data. In this work we solve these issues by introducing a volume deformation method which is real-time, easy to edit with off-the-shelf software and can extrapolate convincingly. To demonstrate the versatility of our method, we apply it in two scenarios: physics-based object deformation and telepresence where avatars are controlled using blendshapes. We also perform thorough experiments showing that our method compares favourably to both volumetric approaches combined with implicit deformation and methods based on mesh deformation.

Abstract— The optics of any camera degrades the sharpness of photographs, which is a key visual quality criterion. This degradation is characterized by the point-spread function (PSF), which depends on the wavelengths of light and is variable across the imaging field. In this paper, we propose a two-step scheme to correct optical aberrations in a single raw or JPEG image, i.e., without any prior information on the camera or lens. First, we estimate local Gaussian blur kernels for overlapping patches and sharpen them with a non-blind deblurring technique. Based on the measurements of the PSFs of dozens of lenses, these blur kernels are modeled as RGB Gaussians defined by seven parameters. Second, we remove the remaining lateral chromatic aberrations (not contemplated in the first step) with a convolutional neural network, trained to minimize the red/green and blue/green residual images. Experiments on both synthetic and real images show that the combination of these two stages yields a fast state-of-the-art blind optical aberration compensation technique that competes with commercial non-blind algorithms.

  • Day 201 (07/20/2022):

Abstract— The ability to reason about changes in the environment is crucial for robots operating over extended periods of time. Agents are expected to capture changes during operation so that actions can be followed to ensure a smooth progression of the working session. However, varying viewing angles and accumulated localization errors make it easy for robots to falsely detect changes in the surrounding world due to low observation overlap and drifted object associations. In this paper, based on the recently proposed category-level Neural Descriptor Fields (NDFs), we develop an object-level online change detection approach that is robust to partially overlapping observations and noisy localization results. Utilizing the shape completion capability and SE(3)-equivariance of NDFs, we represent objects with compact shape codes encoding full object shapes from partial observations. The objects are then organized in a spatial tree structure based on object centers recovered from NDFs for fast queries of object neighborhoods. By associating objects via shape code similarity and comparing local object-neighbor spatial layout, our proposed approach demonstrates robustness to low observation overlap and localization noises. We conduct experiments on both synthetic and real-world sequences and achieve improved change detection results compared to multiple baseline methods. Project webpage: this https URL

Abstract— The FastGeodis package provides an efficient implementation for computing Geodesic and Euclidean distance transforms (or a mixture of both) targeting efficient utilisation of CPU and GPU hardwares. In particular, it implements paralellisable raster scan method from Criminisi et al, where elements in row (2D) or plane (3D) can be computed with parallel threads. This package is able to handle 2D as well as 3D data where it achieves up to 15x speed-up on CPU and up to 60x speed-up on GPU as compared to existing open-source libraries, which uses non-parallelisable single-thread CPU implementation. The performance speed-ups reported here were evaluated using 3D volume data on Nvidia GeForce Titan X (12 GB) with 6-Core Intel Xeon E5-1650 CPU. This package is available at: this https URL

  • Day 202 (07/21/2022):

Abstract— Orchard automation has attracted the attention of researchers recently due to the shortage of global labor force. To automate tasks in orchards such as pruning, thinning, and harvesting, a detailed understanding of the tree structure is required. However, occlusions from foliage and fruits can make it challenging to predict the position of occluded trunks and branches. This work proposes a regression-based deep learning model, Hallucination of Occluded Branch Convolutional Neural Network (HOB-CNN), for tree branch position prediction in varying occluded conditions. We formulate tree branch position prediction as a regression problem towards the horizontal locations of the branch along the vertical direction or vice versa. We present comparative experiments on Y-shaped trees with two state-of-the-art baselines, representing common approaches to the problem. Experiments show that HOB-CNN outperform the baselines at predicting branch position and shows robustness against varying levels of occlusion. We further validated HOB-CNN against two different types of 2D trees, and HOB-CNN shows generalization across different trees and robustness under different occluded conditions.

Abstract— Relations are basic building blocks of human cognition. Classic and recent work suggests that many relations are early developing, and quickly perceived. Machine models that aspire to human-level perception and reasoning should reflect the ability to recognize and reason generatively about relations. We report a systematic empirical examination of a recent text-guided image generation model (DALL-E 2), using a set of 15 basic physical and social relations studied or proposed in the literature, and judgements from human participants (N = 169). Overall, we find that only ~22% of images matched basic relation prompts. Based on a quantitative examination of people's judgments, we suggest that current image generation models do not yet have a grasp of even basic relations involving simple objects and agents. We examine reasons for model successes and failures, and suggest possible improvements based on computations observed in biological intelligence.

  • Day 203 (07/22/2022):

Abstract— Agriculture is the mainstay of human society because it is an essential need for every organism. Paddy cultivation is very significant so far as humans are concerned, largely in the Asian continent, and it is one of the staple foods. However, plant diseases in agriculture lead to depletion in productivity. Plant diseases are generally caused by pests, insects, and pathogens that decrease productivity to a large scale if not controlled within a particular time. Eventually, one cannot see an increase in paddy yield. Accurate and timely identification of plant diseases can help farmers mitigate losses due to pests and diseases. Recently, deep learning techniques have been used to identify paddy diseases and overcome these problems. This paper implements a convolutional neural network (CNN) based on a model and tests a public dataset consisting of 636 infrared image samples with five paddy disease classes and one healthy class. The proposed model proficiently identified and classified paddy diseases of five different types and achieved an accuracy of 88.28%

Abstract— In this work, we address the problem of 4D facial expressions generation. This is usually addressed by animating a neutral 3D face to reach an expression peak, and then get back to the neutral state. In the real world though, people show more complex expressions, and switch from one expression to another. We thus propose a new model that generates transitions between different expressions, and synthesizes long and composed 4D expressions. This involves three sub-problems: (i) modeling the temporal dynamics of expressions, (ii) learning transitions between them, and (iii) deforming a generic mesh. We propose to encode the temporal evolution of expressions using the motion of a set of 3D landmarks, that we learn to generate by training a manifold-valued GAN (Motion3DGAN). To allow the generation of composed expressions, this model accepts two labels encoding the starting and the ending expressions. The final sequence of meshes is generated by a Sparse2Dense mesh Decoder (S2D-Dec) that maps the landmark displacements to a dense, per-vertex displacement of a known mesh topology. By explicitly working with motion trajectories, the model is totally independent from the identity. Extensive experiments on five public datasets show that our proposed approach brings significant improvements with respect to previous solutions, while retaining good generalization to unseen data.

  • Day 204 (07/23/2022):

Abstract— Conventional audio-visual models have independent audio and video branches. We design a unified model for audio and video processing called Unified Audio-Visual Model (UAVM). In this paper, we describe UAVM, report its new state-of-the-art audio-visual event classification accuracy of 65.8% on VGGSound, and describe the intriguing properties of the model.

Abstract— Machine learning and computer vision are dynamically growing fields, which have proven to be able to solve very complex tasks. They could also be used for the monitoring of the honeybee colonies and for the inspection of their health state, which could identify potentially dangerous states before the situation is critical, or to better plan periodic bee colony inspections and therefore save significant costs. In this paper, we present an overview of the state-of-the-art computer vision and machine learning applications used for bee monitoring. We also demonstrate the potential of those methods as an example of an automated bee counter algorithm. The paper is aimed at veterinary and apidology professionals and experts, who might not be familiar with machine learning to introduce to them its possibilities, therefore each family of applications is opened by a brief theoretical introduction and motivation related to its base method. We hope that this paper will inspire other scientists to use the machine learning techniques for other applications in bee monitoring.

  • Day 205 (07/24/2022):

Abstract— Occlusion poses a great threat to monocular multi-person 3D human pose estimation due to large variability in terms of the shape, appearance, and position of occluders. While existing methods try to handle occlusion with pose priors/constraints, data augmentation, or implicit reasoning, they still fail to generalize to unseen poses or occlusion cases and may make large mistakes when multiple people are present. Inspired by the remarkable ability of humans to infer occluded joints from visible cues, we develop a method to explicitly model this process that significantly improves bottom-up multi-person human pose estimation with or without occlusions. First, we split the task into two subtasks: visible keypoints detection and occluded keypoints reasoning, and propose a Deeply Supervised Encoder Distillation (DSED) network to solve the second one. To train our model, we propose a Skeleton-guided human Shape Fitting (SSF) approach to generate pseudo occlusion labels on the existing datasets, enabling explicit occlusion reasoning. Experiments show that explicitly learning from occlusions improves human pose estimation. In addition, exploiting feature-level information of visible joints allows us to reason about occluded joints more accurately. Our method outperforms both the state-of-the-art top-down and bottom-up methods on several benchmarks.

Abstract— Two major causes of death in the United States and worldwide are stroke and myocardial infarction. The underlying cause of both is thrombi released from ruptured or eroded unstable atherosclerotic plaques that occlude vessels in the heart (myocardial infarction) or the brain (stroke). Clinical studies show that plaque composition plays a more important role than lesion size in plaque rupture or erosion events. To determine the plaque composition, various cell types in 3D cardiovascular immunofluorescent images of plaque lesions are counted. However, counting these cells manually is expensive, time-consuming, and prone to human error. These challenges of manual counting motivate the need for an automated approach to localize and count the cells in images. The purpose of this study is to develop an automatic approach to accurately detect and count cells in 3D immunofluorescent images with minimal annotation effort. In this study, we used a weakly supervised learning approach to train the HoVer-Net segmentation model using point annotations to detect nuclei in fluorescent images. The advantage of using point annotations is that they require less effort as opposed to pixel-wise annotation. To train the HoVer-Net model using point annotations, we adopted a popularly used cluster labeling approach to transform point annotations into accurate binary masks of cell nuclei. Traditionally, these approaches have generated binary masks from point annotations, leaving a region around the object unlabeled (which is typically ignored during model training). However, these areas may contain important information that helps determine the boundary between cells. Therefore, we used the entropy minimization loss function in these areas to encourage the model to output more confident predictions on the unlabeled areas. Our comparison studies indicate that the HoVer-Net model trained using our weakly supervised learning approach outperforms baseline methods on the cardiovascular dataset. In addition, we evaluated and compared the performance of the trained HoVer-Net model to other methods on another cardiovascular dataset, which also utilizes DAPI to identify nuclei, but is from a different mouse model stained and imaged independently from the first cardiovascular dataset. The comparison results show the high generalization capability of the HoVer-Net model trained using a weakly supervised learning approach and assessed with standard detection metrics.

  • Day 206 (07/25/2022):

Abstract— We propose a method for estimating the 6DoF pose of a rigid object with an available 3D model from a single RGB image. Unlike classical correspondence-based methods which predict 3D object coordinates at pixels of the input image, the proposed method predicts 3D object coordinates at 3D query points sampled in the camera frustum. The move from pixels to 3D points, which is inspired by recent PIFu-style methods for 3D reconstruction, enables reasoning about the whole object, including its (self-)occluded parts. For a 3D query point associated with a pixel-aligned image feature, we train a fully-connected neural network to predict: (i) the corresponding 3D object coordinates, and (ii) the signed distance to the object surface, with the first defined only for query points in the surface vicinity. We call the mapping realized by this network as Neural Correspondence Field. The object pose is then robustly estimated from the predicted 3D-3D correspondences by the Kabsch-RANSAC algorithm. The proposed method achieves state-of-the-art results on three BOP datasets and is shown superior especially in challenging cases with occlusion. The project website is at: this http URL.

Abstract— Cooperative perception allows a Connected Autonomous Vehicle (CAV) to interact with the other CAVs in the vicinity to enhance perception of surrounding objects to increase safety and reliability. It can compensate for the limitations of the conventional vehicular perception such as blind spots, low resolution, and weather effects. An effective feature fusion model for the intermediate fusion methods of cooperative perception can improve feature selection and information aggregation to further enhance the perception accuracy. We propose adaptive feature fusion models with trainable feature selection modules. One of our proposed models Spatial-wise Adaptive feature Fusion (S-AdaFusion) outperforms all other state-of-the-art models on the two subsets of OPV2V dataset: default CARLA towns for vehicle detection and the Culver City for domain adaptation. In addition, previous studies have only tested cooperative perception for vehicle detection. A pedestrian, however, is much more likely to be seriously injured in a traffic accident. We evaluate the performance of cooperative perception for both vehicle and pedestrian detection using the CODD dataset. Our architecture achieves higher Average Precision (AP) than other existing models for both vehicle and pedestrian detection on the CODD dataset. The experiments demonstrate that cooperative perception also can improve the pedestrian detection accuracy compared to the conventional perception process.

  • Day 207 (07/26/2022):

Abstract— Deep Metric Learning (DML) serves to learn an embedding function to project semantically similar data into nearby embedding space and plays a vital role in many applications, such as image retrieval and face recognition. However, the performance of DML methods often highly depends on sampling methods to choose effective data from the embedding space in the training. In practice, the embeddings in the embedding space are obtained by some deep models, where the embedding space is often with barren area due to the absence of training points, resulting in so called "missing embedding" issue. This issue may impair the sample quality, which leads to degenerated DML performance. In this work, we investigate how to alleviate the "missing embedding" issue to improve the sampling quality and achieve effective DML. To this end, we propose a Densely-Anchored Sampling (DAS) scheme that considers the embedding with corresponding data point as "anchor" and exploits the anchor's nearby embedding space to densely produce embeddings without data points. Specifically, we propose to exploit the embedding space around single anchor with Discriminative Feature Scaling (DFS) and multiple anchors with Memorized Transformation Shifting (MTS). In this way, by combing the embeddings with and without data points, we are able to provide more embeddings to facilitate the sampling process thus boosting the performance of DML. Our method is effortlessly integrated into existing DML frameworks and improves them without bells and whistles. Extensive experiments on three benchmark datasets demonstrate the superiority of our method.

Abstract— The continual appearance of new objects in the visual world poses considerable challenges for current deep learning methods in real-world deployments. The challenge of new task learning is often exacerbated by the scarcity of data for the new categories due to rarity or cost. Here we explore the important task of Few-Shot Class-Incremental Learning (FSCIL) and its extreme data scarcity condition of one-shot. An ideal FSCIL model needs to perform well on all classes, regardless of their presentation order or paucity of data. It also needs to be robust to open-set real-world conditions and be easily adapted to the new tasks that always arise in the field. In this paper, we first reevaluate the current task setting and propose a more comprehensive and practical setting for the FSCIL task. Then, inspired by the similarity of the goals for FSCIL and modern face recognition systems, we propose our method -- Augmented Angular Loss Incremental Classification or ALICE. In ALICE, instead of the commonly used cross-entropy loss, we propose to use the angular penalty loss to obtain well-clustered features. As the obtained features not only need to be compactly clustered but also diverse enough to maintain generalization for future incremental classes, we further discuss how class augmentation, data augmentation, and data balancing affect classification performance. Experiments on benchmark datasets, including CIFAR100, miniImageNet, and CUB200, demonstrate the improved performance of ALICE over the state-of-the-art FSCIL methods.

  • Day 208 (07/27/2022):

Abstract— Video shadow detection aims to generate consistent shadow predictions among video frames. However, the current approaches suffer from inconsistent shadow predictions across frames, especially when the illumination and background textures change in a video. We make an observation that the inconsistent predictions are caused by the shadow feature inconsistency, i.e., the features of the same shadow regions show dissimilar proprieties among the nearby this http URL this paper, we present a novel Shadow-Consistent Correspondence method (SC-Cor) to enhance pixel-wise similarity of the specific shadow regions across frames for video shadow detection. Our proposed SC-Cor has three main advantages. Firstly, without requiring the dense pixel-to-pixel correspondence labels, SC-Cor can learn the pixel-wise correspondence across frames in a weakly-supervised manner. Secondly, SC-Cor considers intra-shadow separability, which is robust to the variant textures and illuminations in videos. Finally, SC-Cor is a plug-and-play module that can be easily integrated into existing shadow detectors with no extra computational cost. We further design a new evaluation metric to evaluate the temporal stability of the video shadow detection results. Experimental results show that SC-Cor outperforms the prior state-of-the-art method, by 6.51% on IoU and 3.35% on the newly introduced temporal stability metric.

Abstract— Monocular depth estimation (MDE) has attracted intense study due to its low cost and critical functions for robotic tasks such as localization, mapping and obstacle detection. Supervised approaches have led to great success with the advance of deep learning, but they rely on large quantities of ground-truth depth annotations that are expensive to acquire. Unsupervised domain adaptation (UDA) transfers knowledge from labeled source data to unlabeled target data, so as to relax the constraint of supervised learning. However, existing UDA approaches may not completely align the domain gap across different datasets because of the domain shift problem. We believe better domain alignment can be achieved via well-designed feature decomposition. In this paper, we propose a novel UDA method for MDE, referred to as Learning Feature Decomposition for Adaptation (LFDA), which learns to decompose the feature space into content and style components. LFDA only attempts to align the content component since it has a smaller domain gap. Meanwhile, it excludes the style component which is specific to the source domain from training the primary task. Furthermore, LFDA uses separate feature distribution estimations to further bridge the domain gap. Extensive experiments on three domain adaptative MDE scenarios show that the proposed method achieves superior accuracy and lower computational cost compared to the state-of-the-art approaches.

  • Day 209 (07/28/2022):

Abstract— In this paper, we propose a novel light field compression method based on a Quantized Distilled Low Rank Neural Radiance Field (QDLR-NeRF) representation. While existing compression methods encode the set of light field sub-aperture images, our proposed method instead learns an implicit scene representation in the form of a Neural Radiance Field (NeRF), which also enables view synthesis. For reducing its size, the model is first learned under a Low Rank (LR) constraint using a Tensor Train (TT) decomposition in an Alternating Direction Method of Multipliers (ADMM) optimization framework. To further reduce the model size, the components of the tensor train decomposition need to be quantized. However, performing the optimization of the NeRF model by simultaneously taking the low rank constraint and the rate-constrained weight quantization into consideration is challenging. To deal with this difficulty, we introduce a network distillation operation that separates the low rank approximation and the weight quantization in the network training. The information from the initial LR constrained NeRF (LR-NeRF) is distilled to a model of a much smaller dimension (DLR-NeRF) based on the TT decomposition of the LR-NeRF. An optimized global codebook is then learned to quantize all TT components, producing the final QDLRNeRF. Experimental results show that our proposed method yields better compression efficiency compared with state-of-the-art methods, and it additionally has the advantage of allowing the synthesis of any light field view with a high quality.

Abstract— Spina Bifida (SB) is a birth defect developed during the early stage of pregnancy in which there is incomplete closing of the spine around the spinal cord. The growing interest in fetoscopic Spina-Bifida repair, which is performed in fetuses who are still in the pregnant uterus, prompts the need for appropriate training. The learning curve for such procedures is steep and requires excellent procedural skills. Computer-based virtual reality (VR) simulation systems offer a safe, cost-effective, and configurable training environment free from ethical and patient safety issues. However, to the best of our knowledge, there are currently no commercial or experimental VR training simulation systems available for fetoscopic SB-repair procedures. In this paper, we propose a novel VR simulator for core manual skills training for SB-repair. An initial simulation realism validation study was carried out by obtaining subjective feedback (face and content validity) from 14 clinicians. The overall simulation realism was on average marked 4.07 on a 5-point Likert scale (1 - very unrealistic, 5 - very realistic). Its usefulness as a training tool for SB-repair as well as in learning fundamental laparoscopic skills was marked 4.63 and 4.80, respectively. These results indicate that VR simulation of fetoscopic procedures may contribute to surgical training without putting fetuses and their mothers at risk. It could also facilitate wider adaptation of fetoscopic procedures in place of much more invasive open fetal surgeries.

  • Day 210 (07/29/2022):

Abstract— Masked autoencoders are scalable vision learners, as the title of MAE \cite{he2022masked}, which suggests that self-supervised learning (SSL) in vision might undertake a similar trajectory as in NLP. Specifically, generative pretext tasks with the masked prediction (e.g., BERT) have become a de facto standard SSL practice in NLP. By contrast, early attempts at generative methods in vision have been buried by their discriminative counterparts (like contrastive learning); however, the success of mask image modeling has revived the masking autoencoder (often termed denoising autoencoder in the past). As a milestone to bridge the gap with BERT in NLP, masked autoencoder has attracted unprecedented attention for SSL in vision and beyond. This work conducts a comprehensive survey of masked autoencoders to shed insight on a promising direction of SSL. As the first to review SSL with masked autoencoders, this work focuses on its application in vision by discussing its historical developments, recent progress, and implications for diverse applications.

Abstract— 3D reconstruction of novel categories based on few-shot learning is appealing in real-world applications and attracts increasing research interests. Previous approaches mainly focus on how to design shape prior models for different categories. Their performance on unseen categories is not very competitive. In this paper, we present a Memory Prior Contrastive Network (MPCN) that can store shape prior knowledge in a few-shot learning based 3D reconstruction framework. With the shape memory, a multi-head attention module is proposed to capture different parts of a candidate shape prior and fuse these parts together to guide 3D reconstruction of novel categories. Besides, we introduce a 3D-aware contrastive learning method, which can not only complement the retrieval accuracy of memory network, but also better organize image features for downstream tasks. Compared with previous few-shot 3D reconstruction methods, MPCN can handle the inter-class variability without category annotations. Experimental results on a benchmark synthetic dataset and the Pascal3D+ real-world dataset show that our model outperforms the current state-of-the-art methods significantly.

  • Day 211 (07/30/2022):

Abstract— Smoke detection has become a significant task in associated industries due to the close relationship between the petrochemical industry's smoke emission and its safety production and environmental damage. There are several production situations in the real industrial production environment, including complete combustion of exhaust gas, inadequate combustion of exhaust gas, direct emission of exhaust gas, etc. We discovered that the datasets used in previous research work can only determine whether smoke is present or not, not its type. That is, the dataset's category does not map to the real-world production situations, which are not conducive to the precise regulation of the production system. As a result, we created a multi-categories smoke detection database that includes a total of 70196 images. We further employed multiple models to conduct the experiment on the proposed database, the results show that the performance of the current algorithms needs to be improved and demonstrate the effectiveness of the proposed database.

Abstract— Most visual retrieval applications store feature vectors for downstream matching tasks. These vectors, from where user information can be spied out, will cause privacy leakage if not carefully protected. To mitigate privacy risks, current works primarily utilize non-invertible transformations or fully cryptographic algorithms. However, transformation-based methods usually fail to achieve satisfying matching performances while cryptosystems suffer from heavy computational overheads. In addition, secure levels of current methods should be improved to confront potential adversary attacks. To address these issues, this paper proposes a plug-in module called SecureVector that protects features by random permutations, 4L-DEC converting and existing homomorphic encryption techniques. For the first time, SecureVector achieves real-time and lossless feature matching among sanitized features, along with much higher security levels than current state-of-the-arts. Extensive experiments on face recognition, person re-identification, image retrieval, and privacy analyses demonstrate the effectiveness of our method. Given limited public projects in this field, codes of our method and implemented baselines are made open-source in this https URL.

  • Day 212 (07/31/2022):

Abstract— Few-shot object detection has been extensively investigated by incorporating meta-learning into region-based detection frameworks. Despite its success, the said paradigm is still constrained by several factors, such as (i) low-quality region proposals for novel classes and (ii) negligence of the inter-class correlation among different classes. Such limitations hinder the generalization of base-class knowledge for the detection of novel-class objects. In this work, we design Meta-DETR, which (i) is the first image-level few-shot detector, and (ii) introduces a novel inter-class correlational meta-learning strategy to capture and leverage the correlation among different classes for robust and accurate few-shot object detection. Meta-DETR works entirely at image level without any region proposals, which circumvents the constraint of inaccurate proposals in prevalent few-shot detection frameworks. In addition, the introduced correlational meta-learning enables Meta-DETR to simultaneously attend to multiple support classes within a single feedforward, which allows to capture the inter-class correlation among different classes, thus significantly reducing the misclassification over similar classes and enhancing knowledge generalization to novel classes. Experiments over multiple few-shot object detection benchmarks show that the proposed Meta-DETR outperforms state-of-the-art methods by large margins. The implementation codes are available at this https URL.

Abstract— LiDAR point clouds, which are usually scanned by rotating LiDAR sensors continuously, capture precise geometry of the surrounding environment and are crucial to many autonomous detection and navigation tasks. Though many 3D deep architectures have been developed, efficient collection and annotation of large amounts of point clouds remain one major challenge in the analytic and understanding of point cloud data. This paper presents PolarMix, a point cloud augmentation technique that is simple and generic but can mitigate the data constraint effectively across different perception tasks and scenarios. PolarMix enriches point cloud distributions and preserves point cloud fidelity via two cross-scan augmentation strategies that cut, edit, and mix point clouds along the scanning direction. The first is scene-level swapping which exchanges point cloud sectors of two LiDAR scans that are cut along the azimuth axis. The second is instance-level rotation and paste which crops point instances from one LiDAR scan, rotates them by multiple angles (to create multiple copies), and paste the rotated point instances into other scans. Extensive experiments show that PolarMix achieves superior performance consistently across different perception tasks and scenarios. In addition, it can work as plug-and-play for various 3D deep architectures and also performs well for unsupervised domain adaptation.

  • Day 213 (08/01/2022):

Abstract— Retrieving the missing dimension information in acoustic images from 2D forward-looking sonar is a well-known problem in the field of underwater robotics. There are works attempting to retrieve 3D information from a single image which allows the robot to generate 3D maps with fly-through motion. However, owing to the unique image formulation principle, estimating 3D information from a single image faces severe ambiguity problems. Classical methods of multi-view stereo can avoid the ambiguity problems, but may require a large number of viewpoints to generate an accurate model. In this work, we propose a novel learning-based multi-view stereo method to estimate 3D information. To better utilize the information from multiple frames, an elevation plane sweeping method is proposed to generate the depth-azimuth-elevation cost volume. The volume after regularization can be considered as a probabilistic volumetric representation of the target. Instead of performing regression on the elevation angles, we use pseudo front depth from the cost volume to represent the 3D information which can avoid the 2D-3D problem in acoustic imaging. High-accuracy results can be generated with only two or three images. Synthetic datasets were generated to simulate various underwater targets. We also built the first real dataset with accurate ground truth in a large scale water tank. Experimental results demonstrate the superiority of our method, compared to other state-of-the-art methods.

Abstract— Category-level object pose estimation aims to predict the 6D pose as well as the 3D metric size of arbitrary objects from a known set of categories. Recent methods harness shape prior adaptation to map the observed point cloud into the canonical space and apply Umeyama algorithm to recover the pose and size. However, their shape prior integration strategy boosts pose estimation indirectly, which leads to insufficient pose-sensitive feature extraction and slow inference speed. To tackle this problem, in this paper, we propose a novel geometry-guided Residual Object Bounding Box Projection network RBP-Pose that jointly predicts object pose and residual vectors describing the displacements from the shape-prior-indicated object surface projections on the bounding box towards the real surface projections. Such definition of residual vectors is inherently zero-mean and relatively small, and explicitly encapsulates spatial cues of the 3D object for robust and accurate pose regression. We enforce geometry-aware consistency terms to align the predicted pose and residual vectors to further boost performance.

  • Day 214 (08/02/2022):

Abstract— Self-supervised learning (SSL) has achieved remarkable performance in pretraining the models that can be further used in downstream tasks via fine-tuning. However, these self-supervised models may not capture meaningful semantic information since the images belonging to the same class are always regarded as negative pairs in the contrastive loss. Consequently, the images of the same class are often located far away from each other in learned feature space, which would inevitably hamper the fine-tuning process. To address this issue, we seek to provide a better initialization for the self-supervised models by enhancing the semantic information. To this end, we propose a Contrastive Initialization (COIN) method that breaks the standard fine-tuning pipeline by introducing an extra initialization stage before fine-tuning. Extensive experiments show that, with the enriched semantics, our COIN significantly outperforms existing methods without introducing extra training cost and sets new state-of-the-arts on multiple downstream tasks.

Abstract— We focus on better understanding the critical factors of augmentation-invariant representation learning. We revisit MoCo v2 and BYOL and try to prove the authenticity of the following assumption: different frameworks bring about representations of different characteristics even with the same pretext task. We establish the first benchmark for fair comparisons between MoCo v2 and BYOL, and observe: (i) sophisticated model configurations enable better adaptation to pre-training dataset; (ii) mismatched optimization strategies of pre-training and fine-tuning hinder model from achieving competitive transfer performances. Given the fair benchmark, we make further investigation and find asymmetry of network structure endows contrastive frameworks to work well under the linear evaluation protocol, while may hurt the transfer performances on long-tailed classification tasks. Moreover, negative samples do not make models more sensible to the choice of data augmentations, nor does the asymmetric network structure. We believe our findings provide useful information for future work.

  • Day 215 (08/03/2022):

Abstract— Neural Radiance Fields (NeRFs) have demonstrated amazing ability to synthesize images of 3D scenes from novel views. However, they rely upon specialized volumetric rendering algorithms based on ray marching that are mismatched to the capabilities of widely deployed graphics hardware. This paper introduces a new NeRF representation based on textured polygons that can synthesize novel images efficiently with standard rendering pipelines. The NeRF is represented as a set of polygons with textures representing binary opacities and feature vectors. Traditional rendering of the polygons with a z-buffer yields an image with features at every pixel, which are interpreted by a small, view-dependent MLP running in a fragment shader to produce a final pixel color. This approach enables NeRFs to be rendered with the traditional polygon rasterization pipeline, which provides massive pixel-level parallelism, achieving interactive frame rates on a wide range of compute platforms, including mobile phones.

Abstract— This paper proposes a 4D backbone for long-term point cloud video understanding. A typical way to capture spatial-temporal context is using 4Dconv or transformer without hierarchy. However, those methods are neither effective nor efficient enough due to camera motion, scene changes, sampling patterns, and the complexity of 4D data. To address those issues, we leverage the primitive plane as a mid-level representation to capture the long-term spatial-temporal context in 4D point cloud videos and propose a novel hierarchical backbone named Point Primitive Transformer(PPTr), which is mainly composed of intra-primitive point transformers and primitive transformers. Extensive experiments show that PPTr outperforms the previous state of the arts on different tasks.

  • Day 216 (08/04/2022):

Abstract— We explore clustering the softmax predictions of deep neural networks and introduce a novel probabilistic clustering method, referred to as k-sBetas. In the general context of clustering distributions, the existing methods focused on exploring distortion measures tailored to simplex data, such as the KL divergence, as alternatives to the standard Euclidean distance. We provide a general perspective of clustering distributions, which emphasizes that the statistical models underlying distortion-based methods may not be descriptive enough. Instead, we optimize a mixed-variable objective measuring the conformity of data within each cluster to the introduced sBeta density function, whose parameters are constrained and estimated jointly with binary assignment variables. Our versatile formulation approximates a variety of parametric densities for modeling cluster data, and enables to control the cluster-balance bias. This yields highly competitive performances for efficient unsupervised adjustment of black-box predictions in a variety of scenarios, including one-shot classification and unsupervised domain adaptation in real-time for road segmentation. Implementation is available at this https URL.

Abstract— Training semantic segmentation models with few annotated samples has great potential in various real-world applications. For the few-shot segmentation task, the main challenge is how to accurately measure the semantic correspondence between the support and query samples with limited training data. To address this problem, we propose to aggregate the learnable covariance matrices with a deformable 4D Transformer to effectively predict the segmentation map. Specifically, in this work, we first devise a novel hard example mining mechanism to learn covariance kernels for the Gaussian process. The learned covariance kernel functions have great advantages over existing cosine similarity-based methods in correspondence measurement. Based on the learned covariance kernels, an efficient doubly deformable 4D Transformer module is designed to adaptively aggregate feature similarity maps into segmentation results. By combining these two designs, the proposed method can not only set new state-of-the-art performance on public benchmarks, but also converge extremely faster than existing methods. Experiments on three public datasets have demonstrated the effectiveness of our method.

  • Day 217 (08/05/2022):

Abstract— In our multicultural world, affect-aware AI systems that support humans need the ability to perceive affect across variations in emotion expression patterns across cultures. These models must perform well in cultural contexts on which they have not been trained. A standard assumption in affective computing is that affect recognition models trained and used within the same culture (intracultural) will perform better than models trained on one culture and used on different cultures (intercultural). We test this assumption and present the first systematic study of intercultural affect recognition models using videos of real-world dyadic interactions from six cultures. We develop an attention-based feature selection approach under temporal causal discovery to identify behavioral cues that can be leveraged in intercultural affect recognition models. Across all six cultures, our findings demonstrate that intercultural affect recognition models were as effective or more effective than intracultural models. We identify and contribute useful behavioral features for intercultural affect recognition; facial features from the visual modality were more useful than the audio modality in this study's context. Our paper presents a proof-of-concept and motivation for the future development of intercultural affect recognition systems.

Abstract— Referring Expression Comprehension (REC) is one of the most important tasks in visual reasoning that requires a model to detect the target object referred by a natural language expression. Among the proposed pipelines, the one-stage Referring Expression Comprehension (OSREC) has become the dominant trend since it merges the region proposal and selection stages. Many state-of-the-art OSREC models adopt a multi-hop reasoning strategy because a sequence of objects is frequently mentioned in a single expression which needs multi-hop reasoning to analyze the semantic relation. However, one unsolved issue of these models is that the number of reasoning steps needs to be pre-defined and fixed before inference, ignoring the varying complexity of expressions. In this paper, we propose a Dynamic Multi-step Reasoning Network, which allows the reasoning steps to be dynamically adjusted based on the reasoning state and expression complexity. Specifically, we adopt a Transformer module to memorize & process the reasoning state and a Reinforcement Learning strategy to dynamically infer the reasoning steps. The work achieves the state-of-the-art performance or significant improvements on several REC datasets, ranging from RefCOCO (+, g) with short expressions, to Ref-Reasoning, a dataset with long and complex compositional expressions.

  • Day 218 (08/06/2022):

Abstract— Graph convolutional network based methods that model the body-joints' relations, have recently shown great promise in 3D skeleton-based human motion prediction. However, these methods have two critical issues: first, deep graph convolutions filter features within only limited graph spectrums, losing sufficient information in the full band; second, using a single graph to model the whole body underestimates the diverse patterns on various body-parts. To address the first issue, we propose adaptive graph scattering, which leverages multiple trainable band-pass graph filters to decompose pose features into richer graph spectrum bands. To address the second issue, body-parts are modeled separately to learn diverse dynamics, which enables finer feature extraction along the spatial dimensions. Integrating the above two designs, we propose a novel skeleton-parted graph scattering network (SPGSN). The cores of the model are cascaded multi-part graph scattering blocks (MPGSBs), building adaptive graph scattering on diverse body-parts, as well as fusing the decomposed features based on the inferred spectrum importance and body-part interactions. Extensive experiments have shown that SPGSN outperforms state-of-the-art methods by remarkable margins of 13.8%, 9.3% and 2.7% in terms of 3D mean per joint position error (MPJPE) on Human3.6M, CMU Mocap and 3DPW datasets, respectively.

Abstract— Neural networks have been rapidly expanding in recent years, with novel strategies and applications. However, challenges such as interpretability, explainability, robustness, safety, trust, and sensibility remain unsolved in neural network technologies, despite the fact that they will unavoidably be addressed for critical applications. Attempts have been made to overcome the challenges in neural network computing by representing and embedding domain knowledge in terms of symbolic representations. Thus, the neuro-symbolic learning (NeSyL) notion emerged, which incorporates aspects of symbolic representation and bringing common sense into neural networks (NeSyL). In domains where interpretability, reasoning, and explainability are crucial, such as video and image captioning, question-answering and reasoning, health informatics, and genomics, NeSyL has shown promising outcomes. This review presents a comprehensive survey on the state-of-the-art NeSyL approaches, their principles, advances in machine and deep learning algorithms, applications such as opthalmology, and most importantly, future perspectives of this emerging field.

  • Day 219 (08/07/2022):

Abstract— Temporal consistency is the key challenge of video depth estimation. Previous works are based on additional optical flow or camera poses, which is time-consuming. By contrast, we derive consistency with less information. Since videos inherently exist with heavy temporal redundancy, a missing frame could be recovered from neighboring ones. Inspired by this, we propose the frame masking network (FMNet), a spatial-temporal transformer network predicting the depth of masked frames based on their neighboring frames. By reconstructing masked temporal features, the FMNet can learn intrinsic inter-frame correlations, which leads to consistency. Compared with prior arts, experimental results demonstrate that our approach achieves comparable spatial accuracy and higher temporal consistency without any additional information. Our work provides a new perspective on consistent video depth estimation. Our official project page is this https URL.

Abstract— Existing metrics used to evaluate table structure recognition algorithms have shortcomings with regard to capturing text and empty cells alignment. In this paper, we build on prior work and propose a new metric - TEDS based IOU similarity (TEDS (IOU)) for table structure recognition which uses bounding boxes instead of text while simultaneously being robust against the above disadvantages. We demonstrate the effectiveness of our metric against previous metrics through various examples.

  • Day 220 (08/08/2022):

Abstract— Introduction: Blood vessels can be non-invasively visualized from a digital fundus image (DFI). Several studies have shown an association between cardiovascular risk and vascular features obtained from DFI. Recent advances in computer vision and image segmentation enable automatising DFI blood vessel segmentation. There is a need for a resource that can automatically compute digital vasculature biomarkers (VBM) from these segmented DFI. Methods: In this paper, we introduce a Python Vasculature BioMarker toolbox, denoted PVBM. A total of 11 VBMs were implemented. In particular, we introduce new algorithmic methods to estimate tortuosity and branching angles. Using PVBM, and as a proof of usability, we analyze geometric vascular differences between glaucomatous patients and healthy controls. Results: We built a fully automated vasculature biomarker toolbox based on DFI segmentations and provided a proof of usability to characterize the vascular changes in glaucoma. For arterioles and venules, all biomarkers were significant and lower in glaucoma patients compared to healthy controls except for tortuosity, venular singularity length and venular branching angles. Conclusion: We have automated the computation of 11 VBMs from retinal blood vessel segmentation. The PVBM toolbox is made open source under a GNU GPL 3 license and is available on this http URL (following publication).

Abstract— Making an accurate prediction of occupancy and flow is essential to enable better safety and interaction for autonomous vehicles under complex traffic scenarios. This work proposes STrajNet: a multi-modal Swin Transformerbased framework for effective scene occupancy and flow predictions. We employ Swin Transformer to encode the image and interaction-aware motion representations and propose a cross-attention module to inject motion awareness into grid cells across different time steps. Flow and occupancy predictions are then decoded through temporalsharing Pyramid decoders. The proposed method shows competitive prediction accuracy and other evaluation metrics in the Waymo Open Dataset benchmark.

  • Day 221 (08/09/2022):

Abstract— Supervised deep learning methods for semantic medical image segmentation are getting increasingly popular in the past few years.However, in resource constrained settings, getting large number of annotated images is very difficult as it mostly requires experts, is expensive and time-consuming.Semi-supervised segmentation can be an attractive solution where a very few labeled images are used along with a large number of unlabeled ones. While the gap between supervised and semi-supervised methods have been dramatically reduced for classification problems in the past couple of years, there still remains a larger gap in segmentation methods. In this work, we adapt a state-of-the-art semi-supervised classification method FixMatch to semantic segmentation task, introducing FixMatchSeg. FixMatchSeg is evaluated in four different publicly available datasets of different anatomy and different modality: cardiac ultrasound, chest X-ray, retinal fundus image, and skin images. When there are few labels, we show that FixMatchSeg performs on par with strong supervised baselines.

Abstract— Recently deep neural networks (DNNs) have achieved significant success in real-world image super-resolution (SR). However, adversarial image samples with quasi-imperceptible noises could threaten deep learning SR models. In this paper, we propose a robust deep learning framework for real-world SR that randomly erases potential adversarial noises in the frequency domain of input images or features. The rationale is that on the SR task clean images or features have a different pattern from the attacked ones in the frequency domain. Observing that existing adversarial attacks usually add high-frequency noises to input images, we introduce a novel random frequency mask module that blocks out high-frequency components possibly containing the harmful perturbations in a stochastic manner. Since the frequency masking may not only destroys the adversarial perturbations but also affects the sharp details in a clean image, we further develop an adversarial sample classifier based on the frequency domain of images to determine if applying the proposed mask module. Based on the above ideas, we devise a novel real-world image SR framework that combines the proposed frequency mask modules and the proposed adversarial classifier with an existing super-resolution backbone network. Experiments show that our proposed method is more insensitive to adversarial attacks and presents more stable SR results than existing models and defenses.

  • Day 222 (08/10/2022):

Abstract— Artistic text recognition is an extremely challenging task with a wide range of applications. However, current scene text recognition methods mainly focus on irregular text while have not explored artistic text specifically. The challenges of artistic text recognition include the various appearance with special-designed fonts and effects, the complex connections and overlaps between characters, and the severe interference from background patterns. To alleviate these problems, we propose to recognize the artistic text at three levels. Firstly, corner points are applied to guide the extraction of local features inside characters, considering the robustness of corner structures to appearance and shape. In this way, the discreteness of the corner points cuts off the connection between characters, and the sparsity of them improves the robustness for background interference. Secondly, we design a character contrastive loss to model the character-level feature, improving the feature representation for character classification. Thirdly, we utilize Transformer to learn the global feature on image-level and model the global relationship of the corner points, with the assistance of a corner-query cross-attention mechanism. Besides, we provide an artistic text dataset to benchmark the performance. Experimental results verify the significant superiority of our proposed method on artistic text recognition and also achieve state-of-the-art performance on several blurred and perspective datasets.

Abstract— Biometric systems represent valid solutions in tasks like user authentication and verification, since they are able to analyze physical and behavioural features with high precision. However, especially when physical biometrics are used, as is the case of iris recognition, they require specific hardware such as retina scanners, sensors, or HD cameras to achieve relevant results. At the same time, they require the users to be very close to the camera to extract high-resolution information. For this reason, in this work, we propose a novel approach that uses long-range (LR) distance images for implementing an iris verification system. More specifically, we present a novel methodology for converting LR iris images into graphs and then use Graph Siamese Neural Networks (GSNN) to predict whether two graphs belong to the same person. In this study, we not only describe this methodology but also evaluate how the spectral components of these images can be used for improving the graph extraction and the final classification task. Results demonstrate the suitability of this approach, encouraging the community to explore graph application in biometric systems.

  • Day 223 (08/11/2022):

Abstract— Generative adversarial networks (GANs) have been trained to be professional artists able to create stunning artworks such as face generation and image style transfer. In this paper, we focus on a realistic business scenario: automated generation of customizable icons given desired mobile applications and theme styles. We first introduce a theme-application icon dataset, termed AppIcon, where each icon has two orthogonal theme and app labels. By investigating a strong baseline StyleGAN2, we observe mode collapse caused by the entanglement of the orthogonal labels. To solve this challenge, we propose IconGAN composed of a conditional generator and dual discriminators with orthogonal augmentations, and a contrastive feature disentanglement strategy is further designed to regularize the feature space of the two discriminators. Compared with other approaches, IconGAN indicates a superior advantage on the AppIcon benchmark. Further analysis also justifies the effectiveness of disentangling app and theme representations. Our project will be released at: this https URL.

Abstract— Applications on Medical Image Analysis suffer from acute shortage of large volume of data properly annotated by medical experts. Supervised Learning algorithms require a large volumes of balanced data to learn robust representations. Often supervised learning algorithms require various techniques to deal with imbalanced data. Self-supervised learning algorithms on the other hand are robust to imbalance in the data and are capable of learning robust representations. In this work, we train a 3D BYOL self-supervised model using gradient accumulation technique to deal with the large number of samples in a batch generally required in a self-supervised algorithm. To the best of our knowledge, this work is one of the first of its kind in this domain. We compare the results obtained through our experiments in the downstream task of ACL Tear Injury detection with the contemporary self-supervised pre-training methods and also with ResNet3D-18 initialized with the Kinetics-400 pre-trained weights. From the downstream task experiments, it is evident that the proposed framework outperforms the existing baselines.

  • Day 224 (08/12/2022):

Abstract— This paper proposes a novel out-of-distribution (OOD) detection framework named MoodCat for image classifiers. MoodCat masks a random portion of the input image and uses a generative model to synthesize the masked image to a new image conditioned on the classification result. It then calculates the semantic difference between the original image and the synthesized one for OOD detection. Compared to existing solutions, MoodCat naturally learns the semantic information of the in-distribution data with the proposed mask and conditional synthesis strategy, which is critical to identifying OODs. Experimental results demonstrate that MoodCat outperforms state-of-the-art OOD detection solutions by a large margin.

Abstract— With the development of generative-based self-supervised learning (SSL) approaches like BeiT and MAE, how to learn good representations by masking random patches of the input image and reconstructing the missing information has grown in concern. However, BeiT and PeCo need a "pre-pretraining" stage to produce discrete codebooks for masked patches representing. MAE does not require a pre-training codebook process, but setting pixels as reconstruction targets may introduce an optimization gap between pre-training and downstream tasks that good reconstruction quality may not always lead to the high descriptive capability for the model. Considering the above issues, in this paper, we propose a simple Self-distillated masked AutoEncoder network, namely SdAE. SdAE consists of a student branch using an encoder-decoder structure to reconstruct the missing information, and a teacher branch producing latent representation of masked tokens. We also analyze how to build good views for the teacher branch to produce latent representation from the perspective of information bottleneck. After that, we propose a multi-fold masking strategy to provide multiple masked views with balanced information for boosting the performance, which can also reduce the computational complexity. Our approach generalizes well: with only 300 epochs pre-training, a vanilla ViT-Base model achieves an 84.1% fine-tuning accuracy on ImageNet-1k classification, 48.6 mIOU on ADE20K segmentation, and 48.9 mAP on COCO detection, which surpasses other methods by a considerable margin. Code is available at this https URL.

  • Day 225 (08/13/2022):

Abstract— As an important upstream task for many medical applications, supervised landmark localization still requires non-negligible annotation costs to achieve desirable performance. Besides, due to cumbersome collection procedures, the limited size of medical landmark datasets impacts the effectiveness of large-scale self-supervised pre-training methods. To address these challenges, we propose a two-stage framework for one-shot medical landmark localization, which first infers landmarks by unsupervised registration from the labeled exemplar to unlabeled targets, and then utilizes these noisy pseudo labels to train robust detectors. To handle the significant structure variations, we learn an end-to-end cascade of global alignment and local deformations, under the guidance of novel loss functions which incorporate edge information. In stage II, we explore self-consistency for selecting reliable pseudo labels and cross-consistency for semi-supervised learning. Our method achieves state-of-the-art performances on public datasets of different body parts, which demonstrates its general applicability.

Abstract— Convolutional neural networks (CNNs) have shown exceptional performance for a range of medical imaging tasks. However, conventional CNNs are not able to explain their reasoning process, therefore limiting their adoption in clinical practice. In this work, we propose an inherently interpretable CNN for regression using similarity-based comparisons (INSightR-Net) and demonstrate our methods on the task of diabetic retinopathy grading. A prototype layer incorporated into the architecture enables visualization of the areas in the image that are most similar to learned prototypes. The final prediction is then intuitively modeled as a mean of prototype labels, weighted by the similarities. We achieved competitive prediction performance with our INSightR-Net compared to a ResNet baseline, showing that it is not necessary to compromise performance for interpretability. Furthermore, we quantified the quality of our explanations using sparsity and diversity, two concepts considered important for a good explanation, and demonstrated the effect of several parameters on the latent space embeddings.

  • Day 226 (08/14/2022):

Abstract— Unlike 2D raster images, there is no single dominant representation for 3D visual data processing. Different formats like point clouds, meshes, or implicit functions each have their strengths and weaknesses. Still, grid representations such as signed distance functions have attractive properties also in 3D. In particular, they offer constant-time random access and are eminently suitable for modern machine learning. Unfortunately, the storage size of a grid grows exponentially with its dimension. Hence they often exceed memory limits even at moderate resolution. This work explores various low-rank tensor formats, including the Tucker, tensor train, and quantics tensor train decompositions, to compress time-varying 3D data. Our method iteratively computes, voxelizes, and compresses each frame's truncated signed distance function and applies tensor rank truncation to condense all frames into a single, compressed tensor that represents the entire 4D scene. We show that low-rank tensor compression is extremely compact to store and query time-varying signed distance functions. It significantly reduces the memory footprint of 4D scenes while surprisingly preserving their geometric quality. Unlike existing iterative learning-based approaches like DeepSDF and NeRF, our method uses a closed-form algorithm with theoretical guarantees.

Abstract— We present a method for transferring the style from a set of images to a 3D object. The texture appearance of an asset is optimized with a differentiable renderer in a pipeline based on losses using pretrained deep neural networks. More specifically, we utilize a nearest-neighbor feature matching loss with CLIP-ResNet50 to extract the style from images. We show that a CLIP- based style loss provides a different appearance over a VGG-based loss by focusing more on texture over geometric shapes. Additionally, we extend the loss to support multiple images and enable loss-based control over the color palette combined with automatic color palette extraction from style images.

  • Day 227 (08/15/2022):

Abstract— Using only a model that was trained to predict where people look at images, and no additional training data, we can produce a range of powerful editing effects for reducing distraction in images. Given an image and a mask specifying the region to edit, we backpropagate through a state-of-the-art saliency model to parameterize a differentiable editing operator, such that the saliency within the masked region is reduced. We demonstrate several operators, including: a recoloring operator, which learns to apply a color transform that camouflages and blends distractors into their surroundings; a warping operator, which warps less salient image regions to cover distractors, gradually collapsing objects into themselves and effectively removing them (an effect akin to inpainting); a GAN operator, which uses a semantic prior to fully replace image regions with plausible, less salient alternatives. The resulting effects are consistent with cognitive research on the human visual system (e.g., since color mismatch is salient, the recoloring operator learns to harmonize objects' colors with their surrounding to reduce their saliency), and, importantly, are all achieved solely through the guidance of the pretrained saliency model, with no additional supervision. We present results on a variety of natural images and conduct a perceptual study to evaluate and validate the changes in viewers' eye-gaze between the original images and our edited results.

Abstract— Human motion modeling is important for many modern graphics applications, which typically require professional skills. In order to remove the skill barriers for laymen, recent motion generation methods can directly generate human motions conditioned on natural languages. However, it remains challenging to achieve diverse and fine-grained motion generation with various text inputs. To address this problem, we propose MotionDiffuse, the first diffusion model-based text-driven motion generation framework, which demonstrates several desired properties over existing methods. 1) Probabilistic Mapping. Instead of a deterministic language-motion mapping, MotionDiffuse generates motions through a series of denoising steps in which variations are injected. 2) Realistic Synthesis. MotionDiffuse excels at modeling complicated data distribution and generating vivid motion sequences. 3) Multi-Level Manipulation. MotionDiffuse responds to fine-grained instructions on body parts, and arbitrary-length motion synthesis with time-varied text prompts. Our experiments show MotionDiffuse outperforms existing SoTA methods by convincing margins on text-driven motion generation and action-conditioned motion generation. A qualitative analysis further demonstrates MotionDiffuse's controllability for comprehensive motion generation. Homepage: this https URL.

  • Day 228 (08/16/2022):

Abstract— There is a constant need for high-performing and computationally efficient neural network models for image super-resolution (SR) often used on low-capacity devices. One way to obtain such models is to compress existing architectures, e.g. quantization. Another option is a neural architecture search (NAS) that discovers new efficient solutions. We propose a novel quantization-aware NAS procedure for a specifically designed SR search space. Our approach performs NAS to find quantization-friendly SR models. The search relies on adding quantization noise to parameters and activations instead of quantizing parameters directly. Our QuantNAS finds architectures with better PSNR/BitOps trade-off than uniform or mixed precision quantization of fixed architectures. Additionally, our search against noise procedure is up to 30% faster than directly quantizing weights.

Abstract— In group activity recognition, hierarchical framework is widely adopted to represent the relationships between individuals and their corresponding group, and has achieved promising performance. However, the existing methods simply employed max/average pooling in this framework, which ignored the distinct contributions of different individuals to the group activity recognition. In this paper, we propose a new contextual pooling scheme, named attentive pooling, which enables the weighted information transition from individual actions to group activity. By utilizing the attention mechanism, the attentive pooling is intrinsically interpretable and able to embed member context into the existing hierarchical model. In order to verify the effectiveness of the proposed scheme, two specific attentive pooling methods, i.e., global attentive pooling (GAP) and hierarchical attentive pooling (HAP) are designed. GAP rewards the individuals that are significant to group activity, while HAP further considers the hierarchical division by introducing subgroup structure. The experimental results on the benchmark dataset demonstrate that our proposal is significantly superior beyond the baseline and is comparable to the state-of-the-art methods.

  • Day 229 (08/17/2022):

Abstract— Temporal Action Localization (TAL) aims to predict both action category and temporal boundary of action instances in untrimmed videos, i.e., start and end time. Fully-supervised solutions are usually adopted in most existing works, and proven to be effective. One of the practical bottlenecks in these solutions is the large amount of labeled training data required. To reduce expensive human label cost, this paper focuses on a rarely investigated yet practical task named semi-supervised TAL and proposes an effective active learning method, named AL-STAL. We leverage four steps for actively selecting video samples with high informativeness and training the localization model, named \emph{Train, Query, Annotate, Append}. Two scoring functions that consider the uncertainty of localization model are equipped in AL-STAL, thus facilitating the video sample rank and selection. One takes entropy of predicted label distribution as measure of uncertainty, named Temporal Proposal Entropy (TPE). And the other introduces a new metric based on mutual information between adjacent action proposals and evaluates the informativeness of video samples, named Temporal Context Inconsistency (TCI). To validate the effectiveness of proposed method, we conduct extensive experiments on two benchmark datasets THUMOS'14 and ActivityNet 1.3. Experiment results show that AL-STAL outperforms the existing competitors and achieves satisfying performance compared with fully-supervised learning.

Abstract— We present a novel method of learning style-agnostic representation using both style transfer and adversarial learning in the reinforcement learning framework. The style, here, refers to task-irrelevant details such as the color of the background in the images, where generalizing the learned policy across environments with different styles is still a challenge. Focusing on learning style-agnostic representations, our method trains the actor with diverse image styles generated from an inherent adversarial style perturbation generator, which plays a min-max game between the actor and the generator, without demanding expert knowledge for data augmentation or additional class labels for adversarial training. We verify that our method achieves competitive or better performances than the state-of-the-art approaches on Procgen and Distracting Control Suite benchmarks, and further investigate the features extracted from our model, showing that the model better captures the invariants and is less distracted by the shifted style. The code is available at this https URL.

  • Day 230 (08/18/2022):

Abstract— To improve the purity of coal and prevent damage to the coal mining machine, it is necessary to identify coal and rock in underground coal mines. At the same time, the mined coal needs to be purified to remove rock and gangue. These two procedures are manually operated by workers in most coal mines. The realization of automatic identification and purification is not only conducive to the automation of coal mines, but also ensures the safety of workers. We discuss the possibility of using image-based methods to distinguish them. In order to find a solution that can be used in both scenarios, a model that forwards image feature extracted by DenseNet to Gaussian process is proposed, which is trained on images taken on surface and achieves high accuracy on images taken underground. This indicates our method is powerful in few-shot learning such as identification of coal and rock/gangue and might be beneficial for realizing automation in coal mines.

Abstract— Unsupervised hashing has attracted much attention for binary representation learning due to the requirement of economical storage and efficiency of binary codes. It aims to encode high-dimensional features in the Hamming space with similarity preservation between instances. However, most existing methods learn hash functions in manifold-based approaches. Those methods capture the local geometric structures (i.e., pairwise relationships) of data, and lack satisfactory performance in dealing with real-world scenarios that produce similar features (e.g. color and shape) with different semantic information. To address this challenge, in this work, we propose an effective unsupervised method, namely Jointly Personalized Sparse Hashing (JPSH), for binary representation learning. To be specific, firstly, we propose a novel personalized hashing module, i.e., Personalized Sparse Hashing (PSH). Different personalized subspaces are constructed to reflect category-specific attributes for different clusters, adaptively mapping instances within the same cluster to the same Hamming space. In addition, we deploy sparse constraints for different personalized subspaces to select important features. We also collect the strengths of the other clusters to build the PSH module with avoiding over-fitting. Then, to simultaneously preserve semantic and pairwise similarities in our JPSH, we incorporate the PSH and manifold-based hash learning into the seamless formulation. As such, JPSH not only distinguishes the instances from different clusters, but also preserves local neighborhood structures within the cluster. Finally, an alternating optimization algorithm is adopted to iteratively capture analytical solutions of the JPSH model. Extensive experiments on four benchmark datasets verify that the JPSH outperforms several hashing algorithms on the similarity search task.

  • Day 231 (08/19/2022):

Abstract— Most unsupervised domain adaptation (UDA) methods assume that labeled source images are available during model adaptation. However, this assumption is often infeasible owing to confidentiality issues or memory constraints on mobile devices. To address these problems, we propose a simple yet effective source-free UDA method that uses only a pre-trained source model and unlabeled target images. Our method captures the aleatoric uncertainty by incorporating data augmentation and trains the feature generator with two consistency objectives. The feature generator is encouraged to learn consistent visual features away from the decision boundaries of the head classifier. Inspired by self-supervised learning, our method promotes inter-space alignment between the prediction space and the feature space while incorporating intra-space consistency within the feature space to reduce the domain gap between the source and target domains. We also consider epistemic uncertainty to boost the model adaptation performance. Extensive experiments on popular UDA benchmarks demonstrate that the performance of our approach is comparable or even superior to vanilla UDA methods without using source images or network modifications.

Abstract— Existing techniques for image-to-image translation commonly have suffered from two critical problems: heavy reliance on per-sample domain annotation and/or inability of handling multiple attributes per image. Recent methods adopt clustering approaches to easily provide per-sample annotations in an unsupervised manner. However, they cannot account for the real-world setting; one sample may have multiple attributes. In addition, the semantics of the clusters are not easily coupled to human understanding. To overcome these, we present a LANguage-driven Image-to-image Translation model, dubbed LANIT. We leverage easy-to-obtain candidate domain annotations given in texts for a dataset and jointly optimize them during training. The target style is specified by aggregating multi-domain style vectors according to the multi-hot domain assignments. As the initial candidate domain texts might be inaccurate, we set the candidate domain texts to be learnable and jointly fine-tune them during training. Furthermore, we introduce a slack domain to cover samples that are not covered by the candidate domains. Experiments on several standard benchmarks demonstrate that LANIT achieves comparable or superior performance to the existing model.

  • Day 232 (08/20/2022):

Abstract— Point cloud registration aims at estimating the geometric transformation between two point cloud scans, in which point-wise correspondence estimation is the key to its success. In addition to previous methods that seek correspondences by hand-crafted or learnt geometric features, recent point cloud registration methods have tried to apply RGB-D data to achieve more accurate correspondence. However, it is not trivial to effectively fuse the geometric and visual information from these two distinctive modalities, especially for the registration problem. In this work, we propose a new Geometry-Aware Visual Feature Extractor (GAVE) that employs multi-scale local linear transformation to progressively fuse these two modalities, where the geometric features from the depth data act as the geometry-dependent convolution kernels to transform the visual features from the RGB data. The resultant visual-geometric features are in canonical feature spaces with alleviated visual dissimilarity caused by geometric changes, by which more reliable correspondence can be achieved. The proposed GAVE module can be readily plugged into recent RGB-D point cloud registration framework. Extensive experiments on 3D Match and ScanNet demonstrate that our method outperforms the state-of-the-art point cloud registration methods even without correspondence or pose supervision. The code is available at: this https URL.

Abstract— The use of Environmental Microorganisms (EMs) offers a highly efficient, low cost and harmless remedy to environmental pollution, by monitoring and decomposing of pollutants. This relies on how the EMs are correctly segmented and identified. With the aim of enhancing the segmentation of weakly visible EM images which are transparent, noisy and have low contrast, a Pairwise Deep Learning Feature Network (PDLF-Net) is proposed in this study. The use of PDLFs enables the network to focus more on the foreground (EMs) by concatenating the pairwise deep learning features of each image to different blocks of the base model SegNet. Leveraging the Shi and Tomas descriptors, we extract each image's deep features on the patches, which are centered at each descriptor using the VGG-16 model. Then, to learn the intermediate characteristics between the descriptors, pairing of the features is performed based on the Delaunay triangulation theorem to form pairwise deep learning features. In this experiment, the PDLF-Net achieves outstanding segmentation results of 89.24%, 63.20%, 77.27%, 35.15%, 89.72%, 91.44% and 89.30% on the accuracy, IoU, Dice, VOE, sensitivity, precision and specificity, respectively.

  • Day 233 (08/21/2022):

Abstract— A considerable amount of research is concerned with the generation of realistic sensor data. LiDAR point clouds are generated by complex simulations or learned generative models. The generated data is usually exploited to enable or improve downstream perception algorithms. Two major questions arise from these procedures: First, how to evaluate the realism of the generated data? Second, does more realistic data also lead to better perception performance? This paper addresses both questions and presents a novel metric to quantify the realism of LiDAR point clouds. Relevant features are learned from real-world and synthetic point clouds by training on a proxy classification task. In a series of experiments, we demonstrate the application of our metric to determine the realism of generated LiDAR data and compare the realism estimation of our metric to the performance of a segmentation model. We confirm that our metric provides an indication for the downstream segmentation performance.

Abstract— A number of private and public insurers compensate workers whose hearing loss can be directly attributed to excessive exposure to noise in the workplace. The claim assessment process is typically lengthy and requires significant effort from human adjudicators who must interpret hand-recorded audiograms, often sent via fax or equivalent. In this work, we present a solution developed in partnership with the Workplace Safety Insurance Board of Ontario to streamline the adjudication process. In particular, we present the first audiogram digitization algorithm capable of automatically extracting the hearing thresholds from a scanned or faxed audiology report as a proof-of-concept. The algorithm extracts most thresholds within 5 dB accuracy, allowing to substantially lessen the time required to convert an audiogram into digital format in a semi-supervised fashion, and is a first step towards the automation of the adjudication process. The source code for the digitization algorithm and a desktop-based implementation of our NIHL annotation portal is publicly available on GitHub (this https URL).

  • Day 234 (08/22/2022):

Abstract— Camera traps, unmanned observation devices, and deep learning-based image recognition systems have greatly reduced human effort in collecting and analyzing wildlife images. However, data collected via above apparatus exhibits 1) long-tailed and 2) open-ended distribution problems. To tackle the open-set long-tailed recognition problem, we propose the Temporal Flow Mask Attention Network that comprises three key building blocks: 1) an optical flow module, 2) an attention residual module, and 3) a meta-embedding classifier. We extract temporal features of sequential frames using the optical flow module and learn informative representation using attention residual blocks. Moreover, we show that applying the meta-embedding technique boosts the performance of the method in open-set long-tailed recognition. We apply this method on a Korean Demilitarized Zone (DMZ) dataset. We conduct extensive experiments, and quantitative and qualitative analyses to prove that our method effectively tackles the open-set long-tailed recognition problem while being robust to unknown classes.

Abstract— Although CLIP-like Visual Language Models provide a functional joint feature space for image and text, due to the limitation of the CILP-like model's image input size (e.g., 224), subtle details are lost in the feature representation if we input high-resolution images (e.g., 2240). In this work, we introduce an efficient framework that can produce a single feature representation for a high-resolution image that injects image details and shares the same semantic space as the original CLIP. In the framework, we train a feature fusing model based on CLIP features extracted from a carefully designed image patch method that can cover objects of any scale, weakly supervised by image-agnostic class prompted queries. We validate our framework by retrieving images from class prompted queries on the real world and synthetic datasets, showing significant performance improvement on these tasks. Furthermore, to fully demonstrate our framework's detail retrieval ability, we construct a CLEVR-like synthetic dataset called CLVER-DS, which is fully annotated and has a controllable object scale.

  • Day 235 (08/23/2022):

Abstract— Image retrieval systems help users to browse and search among extensive images in real-time. With the rise of cloud computing, retrieval tasks are usually outsourced to cloud servers. However, the cloud scenario brings a daunting challenge of privacy protection as cloud servers cannot be fully trusted. To this end, image-encryption-based privacy-preserving image retrieval schemes have been developed, which first extract features from cipher-images, and then build retrieval models based on these features. Yet, most existing approaches extract shallow features and design trivial retrieval models, resulting in insufficient expressiveness for the cipher-images. In this paper, we propose a novel paradigm named Encrypted Vision Transformer (EViT), which advances the discriminative representations capability of cipher-images. First, in order to capture comprehensive ruled information, we extract multi-level local length sequence and global Huffman-code frequency features from the cipher-images which are encrypted by stream cipher during JPEG compression process. Second, we design the Vision Transformer-based retrieval model to couple with the multi-level features, and propose two adaptive data augmentation methods to improve representation power of the retrieval model. Our proposal can be easily adapted to unsupervised and supervised settings via self-supervised contrastive learning manner. Extensive experiments reveal that EViT achieves both excellent encryption and retrieval performance, outperforming current schemes in terms of retrieval accuracy by large margins while protecting image privacy effectively. Code is publicly available at this https URL.

Abstract— Object detection networks have reached an impressive performance level, yet a lack of suitable data in specific applications often limits it in practice. Typically, additional data sources are utilized to support the training task. In these, however, domain gaps between different data sources pose a challenge in deep learning. GAN-based image-to-image style-transfer is commonly applied to shrink the domain gap, but is unstable and decoupled from the object detection task. We propose AWADA, an Attention-Weighted Adversarial Domain Adaptation framework for creating a feedback loop between style-transformation and detection task. By constructing foreground object attention maps from object detector proposals, we focus the transformation on foreground object regions and stabilize style-transfer training. In extensive experiments and ablation studies, we show that AWADA reaches state-of-the-art unsupervised domain adaptation object detection performance in the commonly used benchmarks for tasks such as synthetic-to-real, adverse weather and cross-camera adaptation.

  • Day 236 (08/24/2022):

Abstract— Table structure recognition is a crucial part of document image analysis domain. Its difficulty lies in the need to parse the physical coordinates and logical indices of each cell at the same time. However, the existing methods are difficult to achieve both these goals, especially when the table splitting lines are blurred or tilted. In this paper, we propose an accurate and end-to-end transformer-based table structure recognition method, referred to as TRUST. Transformers are suitable for table structure recognition because of their global computations, perfect memory, and parallel computation. By introducing novel Transformer-based Query-based Splitting Module and Vertex-based Merging Module, the table structure recognition problem is decoupled into two joint optimization sub-tasks: multi-oriented table row/column splitting and table grid merging. The Query-based Splitting Module learns strong context information from long dependencies via Transformer networks, accurately predicts the multi-oriented table row/column separators, and obtains the basic grids of the table accordingly. The Vertex-based Merging Module is capable of aggregating local contextual information between adjacent basic grids, providing the ability to merge basic girds that belong to the same spanning cell accurately. We conduct experiments on several popular benchmarks including PubTabNet and SynthTable, our method achieves new state-of-the-art results. In particular, TRUST runs at 10 FPS on PubTabNet, surpassing the previous methods by a large margin.

Abstract— In order to get raw images of high quality for downstream Image Signal Process (ISP), in this paper we present an Efficient Locally Multiplicative Transformer called ELMformer for raw image restoration. ELMformer contains two core designs especially for raw images whose primitive attribute is single-channel. The first design is a Bi-directional Fusion Projection (BFP) module, where we consider both the color characteristics of raw images and spatial structure of single-channel. The second one is that we propose a Locally Multiplicative Self-Attention (L-MSA) scheme to effectively deliver information from the local space to relevant parts. ELMformer can efficiently reduce the computational consumption and perform well on raw image restoration tasks. Enhanced by these two core designs, ELMformer achieves the highest performance and keeps the lowest FLOPs on raw denoising and raw deblurring benchmarks compared with state-of-the-arts. Extensive experiments demonstrate the superiority and generalization ability of ELMformer. On SIDD benchmark, our method has even better denoising performance than ISP-based methods which need huge amount of additional sRGB training images. The codes are release at this https URL.

  • Day 237 (08/25/2022):

Abstract— Previous unsupervised domain adaptation methods did not handle the cross-domain problem from the perspective of frequency for computer vision. The images or feature maps of different domains can be decomposed into the low-frequency component and high-frequency component. This paper proposes the assumption that low-frequency information is more domain-invariant while the high-frequency information contains domain-related information. Hence, we introduce an approach, named low-frequency module (LFM), to extract domain-invariant feature representations. The LFM is constructed with the digital Gaussian low-pass filter. Our method is easy to implement and introduces no extra hyperparameter. We design two effective ways to utilize the LFM for domain adaptation, and our method is complementary to other existing methods and formulated as a plug-and-play unit that can be combined with these methods. Experimental results demonstrate that our LFM outperforms state-of-the-art methods for various computer vision tasks, including image classification and object detection.

Abstract— 3D object detection from monocular image(s) is a challenging and long-standing problem of computer vision. To combine information from different perspectives without troublesome 2D instance tracking, recent methods tend to aggregate multiview feature by sampling regular 3D grid densely in space, which is inefficient. In this paper, we attempt to improve multi-view feature aggregation by proposing a learnable keypoints sampling method, which scatters pseudo surface points in 3D space, in order to keep data sparsity. The scattered points augmented by multi-view geometric constraints and visual features are then employed to infer objects location and shape in the scene. To make up the limitations of single frame and model multi-view geometry explicitly, we further propose a surface filter module for noise suppression. Experimental results show that our method achieves significantly better performance than previous works in terms of 3D detection (more than 0.1 AP improvement on some categories of ScanNet). The code will be publicly available.

  • Day 238 (08/26/2022):

Abstract— Traditionally, 3D indoor scene reconstruction from posed images happens in two phases: per-image depth estimation, followed by depth merging and surface reconstruction. Recently, a family of methods have emerged that perform reconstruction directly in final 3D volumetric feature space. While these methods have shown impressive reconstruction results, they rely on expensive 3D convolutional layers, limiting their application in resource-constrained environments. In this work, we instead go back to the traditional route, and show how focusing on high quality multi-view depth prediction leads to highly accurate 3D reconstructions using simple off-the-shelf depth fusion. We propose a simple state-of-the-art multi-view depth estimator with two main contributions: 1) a carefully-designed 2D CNN which utilizes strong image priors alongside a plane-sweep feature volume and geometric losses, combined with 2) the integration of keyframe and geometric metadata into the cost volume which allows informed depth plane scoring. Our method achieves a significant lead over the current state-of-the-art for depth estimation and close or better for 3D reconstruction on ScanNet and 7-Scenes, yet still allows for online real-time low-memory reconstruction. Code, models and results are available at this https URL.

Abstract— Smart City applications such as intelligent traffic routing or accident prevention rely on computer vision methods for exact vehicle localization and tracking. Due to the scarcity of accurately labeled data, detecting and tracking vehicles in 3D from multiple cameras proves challenging to explore. We present a massive synthetic dataset for multiple vehicle tracking and segmentation in multiple overlapping and non-overlapping camera views. Unlike existing datasets, which only provide tracking ground truth for 2D bounding boxes, our dataset additionally contains perfect labels for 3D bounding boxes in camera- and world coordinates, depth estimation, and instance, semantic and panoptic segmentation. The dataset consists of 17 hours of labeled video material, recorded from 340 cameras in 64 diverse day, rain, dawn, and night scenes, making it the most extensive dataset for multi-target multi-camera tracking so far. We provide baselines for detection, vehicle re-identification, and single- and multi-camera tracking. Code and data are publicly available.

  • Day 239 (08/27/2022):

Abstract— Capsule networks (CapsNets) aim to parse images into a hierarchical component structure that consists of objects, parts, and their relations. Despite their potential, they are computationally expensive and pose a major drawback, which limits utilizing these networks efficiently on more complex datasets. The current CapsNet models only compare their performance with the capsule baselines and do not perform at the same level as deep CNN-based models on complicated tasks. This paper proposes an efficient way for learning capsules that detect atomic parts of an input image, through a group of SubCapsules, upon which an input vector is projected. Subsequently, we present the Wasserstein Embedding Module that first measures the dissimilarity between the input and components modeled by the SubCapsules, and then finds their degree of alignment based on the learned optimal transport. This strategy leverages new insights on defining alignment between the input and SubCapsules based on the similarity between their respective component distributions. Our proposed model, (i) is lightweight and allows to apply capsules for more complex vision tasks; (ii) performs better than or at par with CNN-based models on these challenging tasks. Our experimental results indicate that Wasserstein Embedding Capsules (WECapsules) perform more robustly on affine transformations, effectively scale up to larger datasets, and outperform the CNN and CapsNet models in several vision tasks.

Abstract— Human motion transfer refers to synthesizing photo-realistic and temporally coherent videos that enable one person to imitate the motion of others. However, current synthetic videos suffer from the temporal inconsistency in sequential frames that significantly degrades the video quality, yet is far from solved by existing methods in the pixel domain. Recently, some works on DeepFake detection try to distinguish the natural and synthetic images in the frequency domain because of the frequency insufficiency of image synthesizing methods. Nonetheless, there is no work to study the temporal inconsistency of synthetic videos from the aspects of the frequency-domain gap between natural and synthetic videos. In this paper, we propose to delve into the frequency space for temporally consistent human motion transfer. First of all, we make the first comprehensive analysis of natural and synthetic videos in the frequency domain to reveal the frequency gap in both the spatial dimension of individual frames and the temporal dimension of the video. To close the frequency gap between the natural and synthetic videos, we propose a novel Frequency-based human MOtion TRansfer framework, named FreMOTR, which can effectively mitigate the spatial artifacts and the temporal inconsistency of the synthesized videos. FreMOTR explores two novel frequency-based regularization modules: 1) the Frequency-domain Appearance Regularization (FAR) to improve the appearance of the person in individual frames and 2) Temporal Frequency Regularization (TFR) to guarantee the temporal consistency between adjacent frames. Finally, comprehensive experiments demonstrate that the FreMOTR not only yields superior performance in temporal consistency metrics but also improves the frame-level visual quality of synthetic videos. In particular, the temporal consistency metrics are improved by nearly 30% than the state-of-the-art model.

  • Day 240 (08/28/2022):

Abstract— The visual quality of point clouds has been greatly emphasized since the ever-increasing 3D vision applications are expected to provide cost-effective and high-quality experiences for users. Looking back on the development of point cloud quality assessment (PCQA) methods, the visual quality is usually evaluated by utilizing single-modal information, i.e., either extracted from the 2D projections or 3D point cloud. The 2D projections contain rich texture and semantic information but are highly dependent on viewpoints, while the 3D point clouds are more sensitive to geometry distortions and invariant to viewpoints. Therefore, to leverage the advantages of both point cloud and projected image modalities, we propose a novel no-reference point cloud quality assessment (NR-PCQA) metric in a multi-modal fashion. In specific, we split the point clouds into sub-models to represent local geometry distortions such as point shift and down-sampling. Then we render the point clouds into 2D image projections for texture feature extraction. To achieve the goals, the sub-models and projected images are encoded with point-based and image-based neural networks. Finally, symmetric cross-modal attention is employed to fuse multi-modal quality-aware information. Experimental results show that our approach outperforms all compared state-of-the-art methods and is far ahead of previous NR-PCQA methods, which highlights the effectiveness of the proposed method.

Abstract— Most existing methods that cope with noisy labels usually assume that the class distributions are well balanced, which has insufficient capacity to deal with the practical scenarios where training samples have imbalanced distributions. To this end, this paper makes an early effort to tackle the image classification task with both long-tailed distribution and label noise. Existing noise-robust learning methods cannot work in this scenario as it is challenging to differentiate noisy samples from clean samples of tail classes. To deal with this problem, we propose a new learning paradigm based on matching between inferences on weak and strong data augmentations to screen out noisy samples and introduce a leave-noise-out regularization to eliminate the effect of the recognized noisy samples. Furthermore, we incorporate a novel prediction penalty based on online prior distribution to avoid bias towards head classes. This mechanism has superiority in capturing the class fitting degree in realtime compared to the existing long-tail classification methods. Exhaustive experiments demonstrate that the proposed method outperforms state-of-the-art algorithms that address the distribution imbalance problem in long-tailed classification under noisy labels.

  • Day 241 (08/29/2022):

Abstract— In this paper, we introduce a new task, spoken video grounding (SVG), which aims to localize the desired video fragments from spoken language descriptions. Compared with using text, employing audio requires the model to directly exploit the useful phonemes and syllables related to the video from raw speech. Moreover, we randomly add environmental noises to this speech audio, further increasing the difficulty of this task and better simulating real applications. To rectify the discriminative phonemes and extract video-related information from noisy audio, we develop a novel video-guided curriculum learning (VGCL) during the audio pre-training process, which can make use of the vital visual perceptions to help understand the spoken language and suppress the external noise. Considering during inference the model can not obtain ground truth video segments, we design a curriculum strategy that gradually shifts the input video from the ground truth to the entire video content during pre-training. Finally, the model can learn how to extract critical visual information from the entire video clip to help understand the spoken language. In addition, we collect the first large-scale spoken video grounding dataset based on ActivityNet, which is named as ActivityNet Speech dataset. Extensive experiments demonstrate our proposed video-guided curriculum learning can facilitate the pre-training process to obtain a mutual audio encoder, significantly promoting the performance of spoken video grounding tasks. Moreover, we prove that in the case of noisy sound, our model outperforms the method that grounding video with ASR transcripts, further demonstrating the effectiveness of our curriculum strategy.

Abstract— Supervised machine learning provides state-of-the-art solutions to a wide range of computer vision problems. However, the need for copious labelled training data limits the capabilities of these algorithms in scenarios where such input is scarce or expensive. Self-supervised learning offers a way to lower the need for manually annotated data by pretraining models for a specific domain on unlabelled data. In this approach, labelled data are solely required to fine-tune models for downstream tasks. Medical image segmentation is a field where labelling data requires expert knowledge and collecting large labelled datasets is challenging; therefore, self-supervised learning algorithms promise substantial improvements in this field. Despite this, self-supervised learning algorithms are used rarely to pretrain medical image segmentation networks. In this paper, we elaborate and analyse the effectiveness of supervised and self-supervised pretraining approaches on downstream medical image segmentation, focusing on convergence and data efficiency. We find that self-supervised pretraining on natural images and target-domain-specific images leads to the fastest and most stable downstream convergence. In our experiments on the ACDC cardiac segmentation dataset, this pretraining approach achieves 4-5 times faster fine-tuning convergence compared to an ImageNet pretrained model. We also show that this approach requires less than five epochs of pretraining on domain-specific data to achieve such improvement in the downstream convergence time. Finally, we find that, in low-data scenarios, supervised ImageNet pretraining achieves the best accuracy, requiring less than 100 annotated samples to realise close to minimal error.

  • Day 242 (08/30/2022):

Abstract— Text-based motion generation models are drawing a surge of interest for their potential for automating the motion-making process in the game, animation, or robot industries. In this paper, we propose a diffusion-based motion synthesis and editing model named FLAME. Inspired by the recent successes in diffusion models, we integrate diffusion-based generative models into the motion domain. FLAME can generate high-fidelity motions well aligned with the given text. Also, it can edit the parts of the motion, both frame-wise and joint-wise, without any fine-tuning. FLAME involves a new transformer-based architecture we devise to better handle motion data, which is found to be crucial to manage variable-length motions and well attend to free-form text. In experiments, we show that FLAME achieves state-of-the-art generation performances on three text-motion datasets: HumanML3D, BABEL, and KIT. We also demonstrate that editing capability of FLAME can be extended to other tasks such as motion prediction or motion in-betweening, which have been previously covered by dedicated models.

Abstract— Existing studies for gait recognition are dominated by in-the-lab scenarios. Since people live in real-world senses, gait recognition in the wild is a more practical problem that has recently attracted the attention of the community of multimedia and computer vision. Current methods that obtain state-of-the-art performance on in-the-lab benchmarks achieve much worse accuracy on the recently proposed in-the-wild datasets because these methods can hardly model the varied temporal dynamics of gait sequences in unconstrained scenes. Therefore, this paper presents a novel multi-hop temporal switch method to achieve effective temporal modeling of gait patterns in real-world scenes. Concretely, we design a novel gait recognition network, named Multi-hop Temporal Switch Network (MTSGait), to learn spatial features and multi-scale temporal features simultaneously. Different from existing methods that use 3D convolutions for temporal modeling, our MTSGait models the temporal dynamics of gait sequences by 2D convolutions. By this means, it achieves high efficiency with fewer model parameters and reduces the difficulty in optimization compared with 3D convolution-based models. Based on the specific design of the 2D convolution kernels, our method can eliminate the misalignment of features among adjacent frames. In addition, a new sampling strategy, i.e., non-cyclic continuous sampling, is proposed to make the model learn more robust temporal features. Finally, the proposed method achieves superior performance on two public gait in-the-wild datasets, i.e., GREW and Gait3D, compared with state-of-the-art methods.

  • Day 243 (08/31/2022):

Abstract— Convolutional Neural Networks (CNNs) are nowadays often employed in vision-based perception stacks for safetycritical applications such as autonomous driving or Unmanned Aerial Vehicles (UAVs). Due to the safety requirements in those use cases, it is important to know the limitations of the CNN and, thus, to detect Out-of-Distribution (OOD) samples. In this work, we present an approach to enable OOD detection for 2D object detection by employing the margin entropy (ME) loss. The proposed method is easy to implement and can be applied to most existing object detection architectures. In addition, we introduce Separability as a metric for detecting OOD samples in object detection. We show that a CNN trained with the ME loss significantly outperforms OOD detection using standard confidence scores. At the same time, the runtime of the underlying object detection framework remains constant rendering the ME loss a powerful tool to enable OOD detection.

Abstract— Holistic scene understanding is pivotal for the performance of autonomous machines. In this paper we propose a new end-to-end model for performing semantic segmentation and depth completion jointly. The vast majority of recent approaches have developed semantic segmentation and depth completion as independent tasks. Our approach relies on RGB and sparse depth as inputs to our model and produces a dense depth map and the corresponding semantic segmentation image. It consists of a feature extractor, a depth completion branch, a semantic segmentation branch and a joint branch which further processes semantic and depth information altogether. The experiments done on Virtual KITTI 2 dataset, demonstrate and provide further evidence, that combining both tasks, semantic segmentation and depth completion, in a multi-task network can effectively improve the performance of each task. Code is available at this https URL depth.

  • Day 244 (09/01/2022):

Abstract— In this paper, we describe a graph-based algorithm that uses the features obtained by a self-supervised transformer to detect and segment salient objects in images and videos. With this approach, the image patches that compose an image or video are organised into a fully connected graph, where the edge between each pair of patches is labeled with a similarity score between patches using features learned by the transformer. Detection and segmentation of salient objects is then formulated as a graph-cut problem and solved using the classical Normalized Cut algorithm. Despite the simplicity of this approach, it achieves state-of-the-art results on several common image and video detection and segmentation tasks. For unsupervised object discovery, this approach outperforms the competing approaches by a margin of 6.1%, 5.7%, and 2.6%, respectively, when tested with the VOC07, VOC12, and COCO20K datasets. For the unsupervised saliency detection task in images, this method improves the score for Intersection over Union (IoU) by 4.4%, 5.6% and 5.2%. When tested with the ECSSD, DUTS, and DUT-OMRON datasets, respectively, compared to current state-of-the-art techniques. This method also achieves competitive results for unsupervised video object segmentation tasks with the DAVIS, SegTV2, and FBMS datasets.

Abstract— Recent works have demonstrated the feasibility of GAN-based morphing attacks that reach similar success rates as more traditional landmark-based methods. This new type of "deep" morphs might require the development of new adequate detectors to protect face recognition systems. We explore simple deep morph detection baselines based on spectral features and LBP histograms features, as well as on CNN models, both in the intra-dataset and cross-dataset case. We observe that simple LBP-based systems are already quite accurate in the intra-dataset setting, but struggle with generalization, a phenomenon that is partially mitigated by fusing together several of those systems at score-level. We conclude that a pretrained ResNet effective for GAN image detection is the most effective overall, reaching close to perfect accuracy. We note however that LBP-based systems maintain a level of interest : additionally to their lower computational requirements and increased interpretability with respect to CNNs, LBP+ResNet fusions sometimes also showcase increased performance versus ResNet-only, hinting that LBP-based systems can focus on meaningful signal that is not necessarily picked up by the CNN detector.

  • Day 245 (09/02/2022):

Abstract— Recognizing human actions from point cloud videos has attracted tremendous attention from both academia and industry due to its wide applications like automatic driving, robotics, and so on. However, current methods for point cloud action recognition usually require a huge amount of data with manual annotations and a complex backbone network with high computation costs, which makes it impractical for real-world applications. Therefore, this paper considers the task of semi-supervised point cloud action recognition. We propose a Masked Pseudo-Labeling autoEncoder (\textbf{MAPLE}) framework to learn effective representations with much fewer annotations for point cloud action recognition. In particular, we design a novel and efficient \textbf{De}coupled \textbf{s}patial-\textbf{t}emporal Trans\textbf{Former} (\textbf{DestFormer}) as the backbone of MAPLE. In DestFormer, the spatial and temporal dimensions of the 4D point cloud videos are decoupled to achieve efficient self-attention for learning both long-term and short-term features. Moreover, to learn discriminative features from fewer annotations, we design a masked pseudo-labeling autoencoder structure to guide the DestFormer to reconstruct features of masked frames from the available frames. More importantly, for unlabeled data, we exploit the pseudo-labels from the classification head as the supervision signal for the reconstruction of features from the masked frames. Finally, comprehensive experiments demonstrate that MAPLE achieves superior results on three public benchmarks and outperforms the state-of-the-art method by 8.08% accuracy on the MSR-Action3D dataset.

Abstract— Human Video Motion Transfer (HVMT) aims to, given an image of a source person, generate his/her video that imitates the motion of the driving person. Existing methods for HVMT mainly exploit Generative Adversarial Networks (GANs) to perform the warping operation based on the flow estimated from the source person image and each driving video frame. However, these methods always generate obvious artifacts due to the dramatic differences in poses, scales, and shifts between the source person and the driving person. To overcome these challenges, this paper presents a novel REgionto-whole human MOtion Transfer (REMOT) framework based on GANs. To generate realistic motions, the REMOT adopts a progressive generation paradigm: it first generates each body part in the driving pose without flow-based warping, then composites all parts into a complete person of the driving motion. Moreover, to preserve the natural global appearance, we design a Global Alignment Module to align the scale and position of the source person with those of the driving person based on their layouts. Furthermore, we propose a Texture Alignment Module to keep each part of the person aligned according to the similarity of the texture. Finally, through extensive quantitative and qualitative experiments, our REMOT achieves state-of-the-art results on two public benchmarks.

  • Day 246 (09/03/2022):

Abstract— We introduce TempCLR, a new time-coherent contrastive learning approach for the structured regression task of 3D hand reconstruction. Unlike previous time-contrastive methods for hand pose estimation, our framework considers temporal consistency in its augmentation scheme, and accounts for the differences of hand poses along the temporal direction. Our data-driven method leverages unlabelled videos and a standard CNN, without relying on synthetic data, pseudo-labels, or specialized architectures. Our approach improves the performance of fully-supervised hand reconstruction methods by 15.9% and 7.6% in PA-V2V on the HO-3D and FreiHAND datasets respectively, thus establishing new state-of-the-art performance. Finally, we demonstrate that our approach produces smoother hand reconstructions through time, and is more robust to heavy occlusions compared to the previous state-of-the-art which we show quantitatively and qualitatively. Our code and models will be available at this https URL.

Abstract— Surface defect detection is one of the most essential processes for industrial quality inspection. Deep learning-based surface defect detection methods have shown great potential. However, the well-performed models usually require large training data and can only detect defects that appeared in the training stage. When facing incremental few-shot data, defect detection models inevitably suffer from catastrophic forgetting and misclassification problem. To solve these problems, this paper proposes a new knowledge distillation network, called Dual Knowledge Align Network (DKAN). The proposed DKAN method follows a pretraining-finetuning transfer learning paradigm and a knowledge distillation framework is designed for fine-tuning. Specifically, an Incremental RCNN is proposed to achieve decoupled stable feature representation of different categories. Under this framework, a Feature Knowledge Align (FKA) loss is designed between class-agnostic feature maps to deal with catastrophic forgetting problems, and a Logit Knowledge Align (LKA) loss is deployed between logit distributions to tackle misclassification problems. Experiments have been conducted on the incremental Few-shot NEU-DET dataset and results show that DKAN outperforms other methods on various few-shot scenes, up to 6.65% on the mean Average Precision metric, which proves the effectiveness of the proposed method.

  • Day 247 (09/04/2022):

Abstract— The point cloud based 3D single object tracking (3DSOT) has drawn increasing attention. Lots of breakthroughs have been made, but we also reveal two severe issues. By an extensive analysis, we find the prediction manner of current approaches is non-robust, i.e., exposing a misalignment gap between prediction score and actually localization accuracy. Another issue is the sparse point returns will damage the feature matching procedure of the SOT task. Based on these insights, we introduce two novel modules, i.e., Adaptive Refine Prediction (ARP) and Target Knowledge Transfer (TKT), to tackle them, respectively. To this end, we first design a strong pipeline to extract discriminative features and conduct the matching procedure with the attention mechanism. Then, ARP module is proposed to tackle the misalignment issue by aggregating all predicted candidates with valuable clues. Finally, TKT module is designed to effectively overcome incomplete point cloud due to sparse and occlusion issues. We call our overall framework PCET. By conducting extensive experiments on the KITTI and Waymo Open Dataset, our model achieves state-of-the-art performance while maintaining a lower computational consumption.

Abstract— Remote sensor image object detection is an important technology for Earth observation, and is used in various tasks such as forest fire monitoring and ocean monitoring. Image object detection technology, despite the significant developments, is struggling to handle remote sensor images and small-scale objects, due to the limited pixels of small objects. Numerous existing studies have demonstrated that an effective way to promote small object detection is to introduce the spatial context. Meanwhile, recent researches for image classification have shown that spectral convolution operations can perceive long-term spatial dependence more efficiently in the frequency domain than spatial domain. Inspired by this observation, we propose a Frequency-aware Feature Pyramid Framework (FFPF) for remote sensing object detection, which consists of a novel Frequency-aware ResNet (F-ResNet) and a Bilateral Spectral-aware Feature Pyramid Network (BS-FPN). Specifically, the F-ResNet is proposed to perceive the spectral context information by plugging the frequency domain convolution into each stage of the backbone, extracting richer features of small objects. To the best of our knowledge, this is the first work to introduce frequency-domain convolution into remote sensing object detection task. In addition, the BSFPN is designed to use a bilateral sampling strategy and skipping connection to better model the association of object features at different scales, towards unleashing the potential of the spectral context information from F-ResNet. Extensive experiments are conducted for object detection in the optical remote sensing image dataset (DIOR and DOTA). The experimental results demonstrate the excellent performance of our method. It achieves an average accuracy (mAP) without any tricks.

  • Day 248 (09/05/2022):

Abstract— This paper addresses the problem of 2D pose representation during unsupervised 2D to 3D pose lifting to improve the accuracy, stability and generalisability of 3D human pose estimation (HPE) models. All unsupervised 2D-3D HPE approaches provide the entire 2D kinematic skeleton to a model during training. We argue that this is sub-optimal and disruptive as long-range correlations are induced between independent 2D key points and predicted 3D ordinates during training. To this end, we conduct the following study. With a maximum architecture capacity of 6 residual blocks, we evaluate the performance of 5 models which each represent a 2D pose differently during the adversarial unsupervised 2D-3D HPE process. Additionally, we show the correlations between 2D key points which are learned during the training process, highlighting the unintuitive correlations induced when an entire 2D pose is provided to a lifting model. Our results show that the most optimal representation of a 2D pose is that of two independent segments, the torso and legs, with no shared features between each lifting network. This approach decreased the average error by 20% on the Human3.6M dataset when compared to a model with a near identical parameter count trained on the entire 2D kinematic skeleton. Furthermore, due to the complex nature of adversarial learning, we show how this representation can also improve convergence during training allowing for an optimum result to be obtained more often.

Abstract— This paper introduces a unified framework for video action segmentation via sequence to sequence (seq2seq) translation in a fully and timestamp supervised setup. In contrast to current state-of-the-art frame-level prediction methods, we view action segmentation as a seq2seq translation task, i.e., mapping a sequence of video frames to a sequence of action segments. Our proposed method involves a series of modifications and auxiliary loss functions on the standard Transformer seq2seq translation model to cope with long input sequences opposed to short output sequences and relatively few videos. We incorporate an auxiliary supervision signal for the encoder via a frame-wise loss and propose a separate alignment decoder for an implicit duration prediction. Finally, we extend our framework to the timestamp supervised setting via our proposed constrained k-medoids algorithm to generate pseudo-segmentations. Our proposed framework performs consistently on both fully and timestamp supervised settings, outperforming or competing state-of-the-art on several datasets.

  • Day 249 (09/06/2022):

Abstract— This paper looks at semi-supervised learning (SSL) for image-based text recognition. One of the most popular SSL approaches is pseudo-labeling (PL). PL approaches assign labels to unlabeled data before re-training the model with a combination of labeled and pseudo-labeled data. However, PL methods are severely degraded by noise and are prone to over-fitting to noisy labels, due to the inclusion of erroneous high confidence pseudo-labels generated from poorly calibrated models, thus, rendering threshold-based selection ineffective. Moreover, the combinatorial complexity of the hypothesis space and the error accumulation due to multiple incorrect autoregressive steps posit pseudo-labeling challenging for sequence models. To this end, we propose a pseudo-label generation and an uncertainty-based data selection framework for semi-supervised text recognition. We first use Beam-Search inference to yield highly probable hypotheses to assign pseudo-labels to the unlabelled examples. Then we adopt an ensemble of models, sampled by applying dropout, to obtain a robust estimate of the uncertainty associated with the prediction, considering both the character-level and word-level predictive distribution to select good quality pseudo-labels. Extensive experiments on several benchmark handwriting and scene-text datasets show that our method outperforms the baseline approaches and the previous state-of-the-art semi-supervised text-recognition methods.

Abstract— In this work, we address the problem of generating speech from silent lip videos for any speaker in the wild. In stark contrast to previous works, our method (i) is not restricted to a fixed number of speakers, (ii) does not explicitly impose constraints on the domain or the vocabulary and (iii) deals with videos that are recorded in the wild as opposed to within laboratory settings. The task presents a host of challenges, with the key one being that many features of the desired target speech, like voice, pitch and linguistic content, cannot be entirely inferred from the silent face video. In order to handle these stochastic variations, we propose a new VAE-GAN architecture that learns to associate the lip and speech sequences amidst the variations. With the help of multiple powerful discriminators that guide the training process, our generator learns to synthesize speech sequences in any voice for the lip movements of any person. Extensive experiments on multiple datasets show that we outperform all baselines by a large margin. Further, our network can be fine-tuned on videos of specific identities to achieve a performance comparable to single-speaker models that are trained on 4× more data. We conduct numerous ablation studies to analyze the effect of different modules of our architecture. We also provide a demo video that demonstrates several qualitative results along with the code and trained models on our website: this http URL.

  • Day 250 (09/07/2022):

Abstract— Event cameras are bio-inspired sensors that perform well in HDR conditions and have high temporal resolution. However, different from traditional frame-based cameras, event cameras measure asynchronous pixel-level brightness changes and return them in a highly discretised format, hence new algorithms are needed. The present paper looks at fronto-parallel motion estimation of an event camera. The flow of the events is modeled by a general homographic warping in a space-time volume, and the objective is formulated as a maximisation of contrast within the image of unwarped events. However, in stark contrast to prior art, we derive a globally optimal solution to this generally non-convex problem, and thus remove the dependency on a good initial guess. Our algorithm relies on branch-and-bound optimisation for which we derive novel, recursive upper and lower bounds for six different contrast estimation functions. The practical validity of our approach is supported by a highly successful application to AGV motion estimation with a downward facing event camera, a challenging scenario in which the sensor experiences fronto-parallel motion in front of noisy, fast moving textures.

Abstract— While previous speech-driven talking face generation methods have made significant progress in improving the visual quality and lip-sync quality of the synthesized videos, they pay less attention to lip motion jitters which greatly undermine the realness of talking face videos. What causes motion jitters, and how to mitigate the problem? In this paper, we conduct systematic analyses on the motion jittering problem based on a state-of-the-art pipeline that uses 3D face representations to bridge the input audio and output video, and improve the motion stability with a series of effective designs. We find that several issues can lead to jitters in synthesized talking face video: 1) jitters from the input 3D face representations; 2) training-inference mismatch; 3) lack of dependency modeling among video frames. Accordingly, we propose three effective solutions to address this issue: 1) we propose a gaussian-based adaptive smoothing module to smooth the 3D face representations to eliminate jitters in the input; 2) we add augmented erosions on the input data of the neural renderer in training to simulate the distortion in inference to reduce mismatch; 3) we develop an audio-fused transformer generator to model dependency among video frames. Besides, considering there is no off-the-shelf metric for measuring motion jitters in talking face video, we devise an objective metric (Motion Stability Index, MSI), to quantitatively measure the motion jitters by calculating the reciprocal of variance acceleration. Extensive experimental results show the superiority of our method on motion-stable face video generation, with better quality than previous systems.

  • Day 251 (09/08/2022):

Abstract— Fully supervised skeleton-based action recognition has achieved great progress with the blooming of deep learning techniques. However, these methods require sufficient labeled data which is not easy to obtain. In contrast, self-supervised skeleton-based action recognition has attracted more attention. With utilizing the unlabeled data, more generalizable features can be learned to alleviate the overfitting problem and reduce the demand of massive labeled training data. Inspired by the MAE, we propose a spatial-temporal masked autoencoder framework for self-supervised 3D skeleton-based action recognition (SkeletonMAE). Following MAE's masking and reconstruction pipeline, we utilize a skeleton based encoder-decoder transformer architecture to reconstruct the masked skeleton sequences. A novel masking strategy, named Spatial-Temporal Masking, is introduced in terms of both joint-level and frame-level for the skeleton sequence. This pre-training strategy makes the encoder output generalizable skeleton features with spatial and temporal dependencies. Given the unmasked skeleton sequence, the encoder is fine-tuned for the action recognition task. Extensive experiments show that our SkeletonMAE achieves remarkable performance and outperforms the state-of-the-art methods on both NTU RGB+D and NTU RGB+D 120 datasets.

Abstract— Scene-text image synthesis techniques aimed at naturally composing text instances on background scene images are very appealing for training deep neural networks because they can provide accurate and comprehensive annotation information. Prior studies have explored generating synthetic text images on two-dimensional and three-dimensional surfaces based on rules derived from real-world observations. Some of these studies have proposed generating scene-text images from learning; however, owing to the absence of a suitable training dataset, unsupervised frameworks have been explored to learn from existing real-world data, which may not result in a robust performance. To ease this dilemma and facilitate research on learning-based scene text synthesis, we propose DecompST, a real-world dataset prepared using public benchmarks, with three types of annotations: quadrilateral-level BBoxes, stroke-level text masks, and text-erased images. Using the DecompST dataset, we propose an image synthesis engine that includes a text location proposal network (TLPNet) and a text appearance adaptation network (TAANet). TLPNet first predicts the suitable regions for text embedding. TAANet then adaptively changes the geometry and color of the text instance according to the context of the background. Our comprehensive experiments verified the effectiveness of the proposed method for generating pretraining data for scene text detectors.

  • Day 252 (09/09/2022):

Abstract— Recognizing soft-biometric pedestrian attributes is essential in video surveillance and fashion retrieval. Recent works show promising results on single datasets. Nevertheless, the generalization ability of these methods under different attribute distributions, viewpoints, varying illumination, and low resolutions remains rarely understood due to strong biases and varying attributes in current datasets. To close this gap and support a systematic investigation, we present UPAR, the Unified Person Attribute Recognition Dataset. It is based on four well-known person attribute recognition datasets: PA100K, PETA, RAPv2, and Market1501. We unify those datasets by providing 3,3M additional annotations to harmonize 40 important binary attributes over 12 attribute categories across the datasets. We thus enable research on generalizable pedestrian attribute recognition as well as attribute-based person retrieval for the first time. Due to the vast variance of the image distribution, pedestrian pose, scale, and occlusion, existing approaches are greatly challenged both in terms of accuracy and efficiency. Furthermore, we develop a strong baseline for PAR and attribute-based person retrieval based on a thorough analysis of regularization methods. Our models achieve state-of-the-art performance in cross-domain and specialization settings on PA100k, PETA, RAPv2, Market1501-Attributes, and UPAR. We believe UPAR and our strong baseline will contribute to the artificial intelligence community and promote research on large-scale, generalizable attribute recognition systems.

Abstract— Long-tailed datasets, where head classes comprise much more training samples than tail classes, cause recognition models to get biased towards the head classes. Weighted loss is one of the most popular ways of mitigating this issue, and a recent work has suggested that class-difficulty might be a better clue than conventionally used class-frequency to decide the distribution of weights. A heuristic formulation was used in the previous work for quantifying the difficulty, but we empirically find that the optimal formulation varies depending on the characteristics of datasets. Therefore, we propose Difficulty-Net, which learns to predict the difficulty of classes using the model's performance in a meta-learning framework. To make it learn reasonable difficulty of a class within the context of other classes, we newly introduce two key concepts, namely the relative difficulty and the driver loss. The former helps Difficulty-Net take other classes into account when calculating difficulty of a class, while the latter is indispensable for guiding the learning to a meaningful direction. Extensive experiments on popular long-tailed datasets demonstrated the effectiveness of the proposed method, and it achieved state-of-the-art performance on multiple long-tailed datasets.

  • Day 253 (09/10/2022):

Abstract— Arbitrary-oriented object detection (AOOD) plays a significant role for image understanding in remote sensing scenarios. The existing AOOD methods face the challenges of ambiguity and high costs in angle representation. To this end, a multi-grained angle representation (MGAR) method, consisting of coarse-grained angle classification (CAC) and fine-grained angle regression (FAR), is proposed. Specifically, the designed CAC avoids the ambiguity of angle prediction by discrete angular encoding (DAE) and reduces complexity by coarsening the granularity of DAE. Based on CAC, FAR is developed to refine the angle prediction with much lower costs than narrowing the granularity of DAE. Furthermore, an Intersection over Union (IoU) aware FAR-Loss (IFL) is designed to improve accuracy of angle prediction using an adaptive re-weighting mechanism guided by IoU. Extensive experiments are performed on several public remote sensing datasets, which demonstrate the effectiveness of the proposed MGAR. Moreover, experiments on embedded devices demonstrate that the proposed MGAR is also friendly for lightweight deployments.

Abstract— Foundation Models (FMs) have demonstrated unprecedented capabilities including zero-shot learning, high fidelity data synthesis, and out of domain generalization. However, as we show in this paper, FMs still have poor out-of-the-box performance on expert tasks (e.g. retrieval of car manuals technical illustrations from language queries), data for which is either unseen or belonging to a long-tail part of the data distribution of the huge datasets used for FM pre-training. This underlines the necessity to explicitly evaluate and finetune FMs on such expert tasks, arguably ones that appear the most in practical real-world applications. In this paper, we propose a first of its kind FETA benchmark built around the task of teaching FMs to understand technical documentation, via learning to match their graphical illustrations to corresponding language descriptions. Our FETA benchmark focuses on text-to-image and image-to-text retrieval in public car manuals and sales catalogue brochures. FETA is equipped with a procedure for completely automatic annotation extraction (code would be released upon acceptance), allowing easy extension of FETA to more documentation types and application domains in the future. Our automatic annotation leads to an automated performance metric shown to be consistent with metrics computed on human-curated annotations (also released). We provide multiple baselines and analysis of popular FMs on FETA leading to several interesting findings that we believe would be very valuable to the FM community, paving the way towards real-world application of FMs for practical expert tasks currently 'overlooked' by standard benchmarks focusing on common objects.

  • Day 254 (09/11/2022):

Abstract— Multi-modal video question answering aims to predict correct answer and localize the temporal boundary relevant to the question. The temporal annotations of questions improve QA performance and interpretability of recent works, but they are usually empirical and costly. To avoid the temporal annotations, we devise a weakly supervised question grounding (WSQG) setting, where only QA annotations are used and the relevant temporal boundaries are generated according to the temporal attention scores. To substitute the temporal annotations, we transform the correspondence between frames and subtitles to Frame-Subtitle (FS) self-supervision, which helps to optimize the temporal attention scores and hence improve the video-language understanding in VideoQA model. The extensive experiments on TVQA and TVQA+ datasets demonstrate that the proposed WSQG strategy gets comparable performance on question grounding, and the FS self-supervision helps improve the question answering and grounding performance on both QA-supervision only and full-supervision settings.

Abstract— Due to the rapid growth of Electrical Capacitance Tomography (ECT) applications in several industrial fields, there is a crucial need for developing high quality, yet fast, methodologies of image reconstruction from raw capacitance measurements. Deep learning, as an effective non-linear mapping tool for complicated functions, has been going viral in many fields including electrical tomography. In this paper, we propose a Conditional Generative Adversarial Network (CGAN) model for reconstructing ECT images from capacitance measurements. The initial image of the CGAN model is constructed from the capacitance measurement. To our knowledge, this is the first time to represent the capacitance measurements in an image form. We have created a new massive ECT dataset of 320K synthetic image measurements pairs for training, and testing the proposed model. The feasibility and generalization ability of the proposed CGAN-ECT model are evaluated using testing dataset, contaminated data and flow patterns that are not exposed to the model during the training phase. The evaluation results prove that the proposed CGAN-ECT model can efficiently create more accurate ECT images than traditional and other deep learning-based image reconstruction algorithms. CGAN-ECT achieved an average image correlation coefficient of more than 99.3% and an average relative image error about 0.07.

  • Day 255 (09/12/2022):

Abstract— Motion-compensated MR reconstruction (MCMR) is a powerful concept with considerable potential, consisting of two coupled sub-problems: Motion estimation, assuming a known image, and image reconstruction, assuming known motion. In this work, we propose a learning-based self-supervised framework for MCMR, to efficiently deal with non-rigid motion corruption in cardiac MR imaging. Contrary to conventional MCMR methods in which the motion is estimated prior to reconstruction and remains unchanged during the iterative optimization process, we introduce a dynamic motion estimation process and embed it into the unrolled optimization. We establish a cardiac motion estimation network that leverages temporal information via a group-wise registration approach, and carry out a joint optimization between the motion estimation and reconstruction. Experiments on 40 acquired 2D cardiac MR CINE datasets demonstrate that the proposed unrolled MCMR framework can reconstruct high quality MR images at high acceleration rates where other state-of-the-art methods fail. We also show that the joint optimization mechanism is mutually beneficial for both sub-tasks, i.e., motion estimation and image reconstruction, especially when the MR image is highly undersampled.

Abstract— Diffusion models are a class of deep generative models that have shown impressive results on various tasks with a solid theoretical foundation. Despite demonstrated success than state-of-the-art approaches, diffusion models often entail costly sampling procedures and sub-optimal likelihood estimation. Significant efforts have been made to improve the performance of diffusion models in various aspects. In this article, we present a comprehensive review of existing variants of diffusion models. Specifically, we provide the taxonomy of diffusion models and categorize them into three types: sampling-acceleration enhancement, likelihood-maximization enhancement, and data-generalization enhancement. We also introduce the other generative models (i.e., variational autoencoders, generative adversarial networks, normalizing flow, autoregressive models, and energy-based models) and discuss the connections between diffusion models and these generative models. Then we review the applications of diffusion models, including computer vision, natural language processing, waveform signal processing, multi-modal modeling, molecular graph generation, time series modeling, and adversarial purification. Furthermore, we propose new perspectives pertaining to the development of generative models. Github: this https URL.

  • Day 256 (09/13/2022):

Abstract— Masked autoencoders have become popular training paradigms for self-supervised visual representation learning. These models randomly mask a portion of the input and reconstruct the masked portion according to the target representations. In this paper, we first show that a careful choice of the target representation is unnecessary for learning good representations, since different targets tend to derive similarly behaved models. Driven by this observation, we propose a multi-stage masked distillation pipeline and use a randomly initialized model as the teacher, enabling us to effectively train high-capacity models without any efforts to carefully design target representations. Interestingly, we further explore using teachers of larger capacity, obtaining distilled students with remarkable transferring ability. On different tasks of classification, transfer learning, object detection, and semantic segmentation, the proposed method to perform masked knowledge distillation with bootstrapped teachers (dBOT) outperforms previous self-supervised methods by nontrivial margins. We hope our findings, as well as the proposed method, could motivate people to rethink the roles of target representations in pre-training masked autoencoders.

Abstract— Synthetic aperture sonar (SAS) imagery is crucial for several applications, including target recognition and environmental segmentation. Deep learning models have led to much success in SAS analysis; however, the features extracted by these approaches may not be suitable for capturing certain textural information. To address this problem, we present a novel application of histogram layers on SAS imagery. The addition of histogram layer(s) within the deep learning models improved performance by incorporating statistical texture information on both synthetic and real-world datasets.

  • Day 257 (09/14/2022):

Abstract— In this paper, we introduce novel lightweight generative adversarial networks, which can effectively capture long-range dependencies in the image generation process, and produce high-quality results with a much simpler architecture. To achieve this, we first introduce a long-range module, allowing the network to dynamically adjust the number of focused sampling pixels and to also augment sampling locations. Thus, it can break the limitation of the fixed geometric structure of the convolution operator, and capture long-range dependencies in both spatial and channel-wise directions. Also, the proposed long-range module can highlight negative relations between pixels, working as a regularization to stabilize training. Furthermore, we propose a new generation strategy through which we introduce metadata into the image generation process to provide basic information about target images, which can stabilize and speed up the training process. Our novel long-range module only introduces few additional parameters and is easily inserted into existing models to capture long-range dependencies. Extensive experiments demonstrate the competitive performance of our method with a lightweight architecture.

Abstract— Adversarial training (AT) with samples generated by Fast Gradient Sign Method (FGSM), also known as FGSM-AT, is a computationally simple method to train robust networks. However, during its training procedure, an unstable mode of "catastrophic overfitting" has been identified in arXiv:2001.03994 [cs.LG], where the robust accuracy abruptly drops to zero within a single training step. Existing methods use gradient regularizers or random initialization tricks to attenuate this issue, whereas they either take high computational cost or lead to lower robust accuracy. In this work, we provide the first study, which thoroughly examines a collection of tricks from three perspectives: Data Initialization, Network Structure, and Optimization, to overcome the catastrophic overfitting in FGSM-AT. Surprisingly, we find that simple tricks, i.e., a) masking partial pixels (even without randomness), b) setting a large convolution stride and smooth activation functions, or c) regularizing the weights of the first convolutional layer, can effectively tackle the overfitting issue. Extensive results on a range of network architectures validate the effectiveness of each proposed trick, and the combinations of tricks are also investigated. For example, trained with PreActResNet-18 on CIFAR-10, our method attains 49.8% accuracy against PGD-50 attacker and 46.4% accuracy against AutoAttack, demonstrating that pure FGSM-AT is capable of enabling robust learners. The code and models are publicly available at this https URL.

  • Day 258 (09/15/2022):

Abstract— Adversarial training (AT) with samples generated by Fast Gradient Sign Method (FGSM), also known as FGSM-AT, is a computationally simple method to train robust networks. However, during its training procedure, an unstable mode of "catastrophic overfitting" has been identified in arXiv:2001.03994 [cs.LG], where the robust accuracy abruptly drops to zero within a single training step. Existing methods use gradient regularizers or random initialization tricks to attenuate this issue, whereas they either take high computational cost or lead to lower robust accuracy. In this work, we provide the first study, which thoroughly examines a collection of tricks from three perspectives: Data Initialization, Network Structure, and Optimization, to overcome the catastrophic overfitting in FGSM-AT. Surprisingly, we find that simple tricks, i.e., a) masking partial pixels (even without randomness), b) setting a large convolution stride and smooth activation functions, or c) regularizing the weights of the first convolutional layer, can effectively tackle the overfitting issue. Extensive results on a range of network architectures validate the effectiveness of each proposed trick, and the combinations of tricks are also investigated. For example, trained with PreActResNet-18 on CIFAR-10, our method attains 49.8% accuracy against PGD-50 attacker and 46.4% accuracy against AutoAttack, demonstrating that pure FGSM-AT is capable of enabling robust learners. The code and models are publicly available at this https URL.

Abstract— In video analysis, background models have many applications such as background/foreground separation, change detection, anomaly detection, tracking, and more. However, while learning such a model in a video captured by a static camera is a fairly-solved task, in the case of a Moving-camera Background Model (MCBM), the success has been far more modest due to algorithmic and scalability challenges that arise due to the camera motion. Thus, existing MCBMs are limited in their scope and their supported camera-motion types. These hurdles also impeded the employment, in this unsupervised task, of end-to-end solutions based on deep learning (DL). Moreover, existing MCBMs usually model the background either on the domain of a typically-large panoramic image or in an online fashion. Unfortunately, the former creates several problems, including poor scalability, while the latter prevents the recognition and leveraging of cases where the camera revisits previously-seen parts of the scene. This paper proposes a new method, called DeepMCBM, that eliminates all the aforementioned issues and achieves state-of-the-art results. Concretely, first we identify the difficulties associated with joint alignment of video frames in general and in a DL setting in particular. Next, we propose a new strategy for joint alignment that lets us use a spatial transformer net with neither a regularization nor any form of specialized (and non-differentiable) initialization. Coupled with an autoencoder conditioned on unwarped robust central moments (obtained from the joint alignment), this yields an end-to-end regularization-free MCBM that supports a broad range of camera motions and scales gracefully. We demonstrate DeepMCBM's utility on a variety of videos, including ones beyond the scope of other methods. Our code is available at this https URL.

  • Day 259 (09/16/2022):

317 DPFNet: A Dual-branch Dilated Network with Phase-aware Fourier Convolution for Low-light Image Enhancement

Abstract— Low-light image enhancement is a classical computer vision problem aiming to recover normal-exposure images from low-light images. However, convolutional neural networks commonly used in this field are good at sampling low-frequency local structural features in the spatial domain, which leads to unclear texture details of the reconstructed images. To alleviate this problem, we propose a novel module using the Fourier coefficients, which can recover high-quality texture details under the constraint of semantics in the frequency phase and supplement the spatial domain. In addition, we design a simple and efficient module for the image spatial domain using dilated convolutions with different receptive fields to alleviate the loss of detail caused by frequent downsampling. We integrate the above parts into an end-to-end dual branch network and design a novel loss committee and an adaptive fusion module to guide the network to flexibly combine spatial and frequency domain features to generate more pleasing visual effects. Finally, we evaluate the proposed network on public benchmarks. Extensive experimental results show that our method outperforms many existing state-of-the-art ones, showing outstanding performance and potential.

318 Traffic Congestion Prediction using Deep Convolutional Neural Networks: A Color-Coding Approach

Abstract— The traffic video data has become a critical factor in confining the state of traffic congestion due to the recent advancements in computer vision. This work proposes a unique technique for traffic video classification using a color-coding scheme before training the traffic data in a Deep convolutional neural network. At first, the video data is transformed into an imagery data set; then, the vehicle detection is performed using the You Only Look Once algorithm. A color-coded scheme has been adopted to transform the imagery dataset into a binary image dataset. These binary images are fed to a Deep Convolutional Neural Network. Using the UCSD dataset, we have obtained a classification accuracy of 98.2%.

  • Day 260 (09/17/2022):

319 Omni-Dimensional Dynamic Convolution

Abstract— Learning a single static convolutional kernel in each convolutional layer is the common training paradigm of modern Convolutional Neural Networks (CNNs). Instead, recent research in dynamic convolution shows that learning a linear combination of n convolutional kernels weighted with their input-dependent attentions can significantly improve the accuracy of light-weight CNNs, while maintaining efficient inference. However, we observe that existing works endow convolutional kernels with the dynamic property through one dimension (regarding the convolutional kernel number) of the kernel space, but the other three dimensions (regarding the spatial size, the input channel number and the output channel number for each convolutional kernel) are overlooked. Inspired by this, we present Omni-dimensional Dynamic Convolution (ODConv), a more generalized yet elegant dynamic convolution design, to advance this line of research. ODConv leverages a novel multi-dimensional attention mechanism with a parallel strategy to learn complementary attentions for convolutional kernels along all four dimensions of the kernel space at any convolutional layer. As a drop-in replacement of regular convolutions, ODConv can be plugged into many CNN architectures. Extensive experiments on the ImageNet and MS-COCO datasets show that ODConv brings solid accuracy boosts for various prevailing CNN backbones including both light-weight and large ones, e.g., 3.77%~5.71%|1.86%~3.72% absolute top-1 improvements to MobivleNetV2|ResNet family on the ImageNet dataset. Intriguingly, thanks to its improved feature learning ability, ODConv with even one single kernel can compete with or outperform existing dynamic convolution counterparts with multiple kernels, substantially reducing extra parameters. Furthermore, ODConv is also superior to other attention modules for modulating the output features or the convolutional weights.

320 On Developing Facial Stress Analysis and Expression Recognition Platform

Abstract— This work represents the experimental and development process of system facial expression recognition and facial stress analysis algorithms for an immersive digital learning platform. The system retrieves from users web camera and evaluates it using artificial neural network (ANN) algorithms. The ANN output signals can be used to score and improve the learning process. Adapting an ANN to a new system can require a significant implementation effort or the need to repeat the ANN training. There are also limitations related to the minimum hardware required to run an ANN. To overpass these constraints, some possible implementations of facial expression recognition and facial stress analysis algorithms in real-time systems are presented. The implementation of the new solution has made it possible to improve the accuracy in the recognition of facial expressions and also to increase their response speed. Experimental results showed that using the developed algorithms allow to detect the heart rate with better rate in comparison with social equipment.

  • Day 261 (09/18/2022):

321 An Attention-Guided Multistream Feature Fusion Network for Localization of Risky Objects in Driving Videos

Abstract— Detecting dangerous traffic agents in videos captured by vehicle-mounted dashboard cameras (dashcams) is essential to facilitate safe navigation in a complex environment. Accident-related videos are just a minor portion of the driving video big data, and the transient pre-accident processes are highly dynamic and complex. Besides, risky and non-risky traffic agents can be similar in their appearance. These make risky object localization in the driving video particularly challenging. To this end, this paper proposes an attention-guided multistream feature fusion network (AM-Net) to localize dangerous traffic agents from dashcam videos. Two Gated Recurrent Unit (GRU) networks use object bounding box and optical flow features extracted from consecutive video frames to capture spatio-temporal cues for distinguishing dangerous traffic agents. An attention module coupled with the GRUs learns to attend to the traffic agents relevant to an accident. Fusing the two streams of features, AM-Net predicts the riskiness scores of traffic agents in the video. In supporting this study, the paper also introduces a benchmark dataset called Risky Object Localization (ROL). The dataset contains spatial, temporal, and categorical annotations with the accident, object, and scene-level attributes. The proposed AM-Net achieves a promising performance of 85.73% AUC on the ROL dataset. Meanwhile, the AM-Net outperforms current state-of-the-art for video anomaly detection by 6.3% AUC on the DoTA dataset. A thorough ablation study further reveals AM-Net's merits by evaluating the contributions of its different components.

322 StyleGAN Encoder-based Attack for Block Scrambled Face Images

Abstract— In this paper, we propose an attack method to block scrambled face images, particularly Encryption-then-Compression (EtC) applied images by utilizing the existing powerful StyleGAN encoder and decoder for the first time. Instead of reconstructing identical images as plain ones from encrypted images, we focus on recovering styles that can reveal identifiable information from the encrypted images. The proposed method trains an encoder by using plain and encrypted image pairs with a particular training strategy. While state-of-the-art attack methods cannot recover any perceptual information from EtC images, the proposed method discloses personally identifiable information such as hair color, skin color, eyeglasses, gender, etc. Experiments were carried out on the CelebA dataset, and results show that reconstructed images have some perceptual similarities compared to plain images.

  • Day 262 (09/19/2022):

323 Towards Bridging the Performance Gaps of Joint-Energy-based Models

Abstract— Can we train a hybrid discriminative-generative model within a single network? This question has recently been answered in the affirmative, introducing the field of Joint Energy-based Model (JEM), which achieves high classification accuracy and image generation quality simultaneously. Despite recent advances, there remain two performance gaps: the accuracy gap to the standard softmax classifier, and the generation quality gap to state-of-the-art generative models. In this paper, we introduce a variety of training techniques to bridge the accuracy gap and the generation quality gap of JEM. 1) We incorporate a recently proposed sharpness-aware minimization (SAM) framework to train JEM, which promotes the energy landscape smoothness and the generalizability of JEM. 2) We exclude data augmentation from the maximum likelihood estimate pipeline of JEM, and mitigate the negative impact of data augmentation to image generation quality. Extensive experiments on multiple datasets demonstrate that our SADA-JEM achieves state-of-the-art performances and outperforms JEM in image classification, image generation, calibration, out-of-distribution detection and adversarial robustness by a notable margin.

324 Causes of Catastrophic Forgetting in Class-Incremental Semantic Segmentation

Abstract— Class-incremental learning for semantic segmentation (CiSS) is presently a highly researched field which aims at updating a semantic segmentation model by sequentially learning new semantic classes. A major challenge in CiSS is overcoming the effects of catastrophic forgetting, which describes the sudden drop of accuracy on previously learned classes after the model is trained on a new set of classes. Despite latest advances in mitigating catastrophic forgetting, the underlying causes of forgetting specifically in CiSS are not well understood. Therefore, in a set of experiments and representational analyses, we demonstrate that the semantic shift of the background class and a bias towards new classes are the major causes of forgetting in CiSS. Furthermore, we show that both causes mostly manifest themselves in deeper classification layers of the network, while the early layers of the model are not affected. Finally, we demonstrate how both causes are effectively mitigated utilizing the information contained in the background, with the help of knowledge distillation and an unbiased cross-entropy loss.

  • Day 263 (09/20/2022):

325 Continual Learning for Class- and Domain-Incremental Semantic Segmentation

Abstract— The field of continual deep learning is an emerging field and a lot of progress has been made. However, concurrently most of the approaches are only tested on the task of image classification, which is not relevant in the field of intelligent vehicles. Only recently approaches for class-incremental semantic segmentation were proposed. However, all of those approaches are based on some form of knowledge distillation. At the moment there are no investigations on replay-based approaches that are commonly used for object recognition in a continual setting. At the same time while unsupervised domain adaption for semantic segmentation gained a lot of traction, investigations regarding domain-incremental learning in an continual setting is not well-studied. Therefore, the goal of our work is to evaluate and adapt established solutions for continual object recognition to the task of semantic segmentation and to provide baseline methods and evaluation protocols for the task of continual semantic segmentation. We firstly introduce evaluation protocols for the class- and domain-incremental segmentation and analyze selected approaches. We show that the nature of the task of semantic segmentation changes which methods are most effective in mitigating forgetting compared to image classification. Especially, in class-incremental learning knowledge distillation proves to be a vital tool, whereas in domain-incremental learning replay methods are the most effective method.

326 Spatial-then-Temporal Self-Supervised Learning for Video Correspondence

Abstract— Learning temporal correspondence from unlabeled videos is of vital importance in computer vision, and has been tackled by different kinds of self-supervised pretext tasks. For the self-supervised learning, recent studies suggest using large-scale video datasets despite the training cost. We propose a spatial-then-temporal pretext task to address the training data cost problem. The task consists of two steps. First, we use contrastive learning from unlabeled still image data to obtain appearance-sensitive features. Then we switch to unlabeled video data and learn motion-sensitive features by reconstructing frames. In the second step, we propose a global correlation distillation loss to retain the appearance sensitivity learned in the first step, as well as a local correlation distillation loss in a pyramid structure to combat temporal discontinuity. Experimental results demonstrate that our method surpasses the state-of-the-art self-supervised methods on a series of correspondence-based tasks. The conducted ablation studies verify the effectiveness of the proposed two-step task and loss functions.

  • Day 264 (09/21/2022):

327 A Large-Scale Multiple-objective Method for Black-Box Attack Against Object Detection

Abstract— Recent studies have shown that detectors based on deep models are vulnerable to adversarial examples, even in the black-box scenario where the attacker cannot access the model information. Most existing attack methods aim to minimize the true positive rate, which often shows poor attack performance, as another sub-optimal bounding box may be detected around the attacked bounding box to be the new true positive one. To settle this challenge, we propose to minimize the true positive rate and maximize the false positive rate, which can encourage more false positive objects to block the generation of new true positive bounding boxes. It is modeled as a multi-objective optimization (MOP) problem, of which the generic algorithm can search the Pareto-optimal. However, our task has more than two million decision variables, leading to low searching efficiency. Thus, we extend the standard Genetic Algorithm with Random Subset selection and Divide-and-Conquer, called GARSDC, which significantly improves the efficiency. Moreover, to alleviate the sensitivity to population quality in generic algorithms, we generate a gradient-prior initial population, utilizing the transferability between different detectors with similar backbones. Compared with the state-of-art attack methods, GARSDC decreases by an average 12.0 in the mAP and queries by about 1000 times in extensive experiments. Our codes can be found at this https URL GARSDC.

328 KaliCalib: A Framework for Basketball Court Registration

Abstract— Tracking the players and the ball in team sports is key to analyse the performance or to enhance the game watching experience with augmented reality. When the only sources for this data are broadcast videos, sports-field registration systems are required to estimate the homography and re-project the ball or the players from the image space to the field space. This paper describes a new basketball court registration framework in the context of the MMSports 2022 camera calibration challenge. The method is based on the estimation by an encoder-decoder network of the positions of keypoints sampled with perspective-aware constraints. The regression of the basket positions and heavy data augmentation techniques make the model robust to different arenas. Ablation studies show the positive effects of our contributions on the challenge test set. Our method divides the mean squared error by 4.7 compared to the challenge baseline.

  • Day 265 (09/22/2022):

329 TwistSLAM++: Fusing Multiple Modalities for accurate dynamic semantic SLAM

Abstract— Most classical SLAM systems rely on the static scene assumption, which limits their applicability in real world scenarios. Recent SLAM frameworks have been proposed to simultaneously track the camera and moving objects. However they are often unable to estimate the canonical pose of the objects and exhibit a low object tracking accuracy. To solve this problem we propose TwistSLAM++, a semantic, dynamic, SLAM system that fuses stereo images and LiDAR information. Using semantic information, we track potentially moving objects and associate them to 3D object detections in LiDAR scans to obtain their pose and size. Then, we perform registration on consecutive object scans to refine object pose estimation. Finally, object scans are used to estimate the shape of the object and constrain map points to lie on the estimated surface within the BA. We show on classical benchmarks that this fusion approach based on multimodal information improves the accuracy of object tracking.

330 3DPointCaps++: Learning 3D Representations with Capsule Networks

Abstract— We present 3DPointCaps++ for learning robust, flexible and generalizable 3D object representations without requiring heavy annotation efforts or supervision. Unlike conventional 3D generative models, our algorithm aims for building a structured latent space where certain factors of shape variations, such as object parts, can be disentangled into independent sub-spaces. Our novel decoder then acts on these individual latent sub-spaces (i.e. capsules) using deconvolution operators to reconstruct 3D points in a self-supervised manner. We further introduce a cluster loss ensuring that the points reconstructed by a single capsule remain local and do not spread across the object uncontrollably. These contributions allow our network to tackle the challenging tasks of part segmentation, part interpolation/replacement as well as correspondence estimation across rigid / non-rigid shape, and across / within category. Our extensive evaluations on ShapeNet objects and human scans demonstrate that our network can learn generic representations that are robust and useful in many applications.

  • Day 266 (09/23/2022):

331 A Trio-Method for Retinal Vessel Segmentation using Image Processing

Abstract— Inner Retinal neurons are a most essential part of th

About

This is a repo where I keep a track of the Computer Vision papers that I am reading

computervisionmachinelearningmachinelearning-pythonpaperresearch