Learning useful representations with the unsupervised or weakly supervised methods is a key challenge in artificial intelligence. The representation learning (RL) is helpful for most specific tasks like classification, recognition, detection, image editing, image retrieval, et al. in computer vision area. RL is mainly appeared in the learning good representations for down-stream task, disentangled representation/attributes, VAE, GAN, Flow-based model, image-translation, deep clustering papers. This repo mainly focuses on the lasted development in the RL area.
I hope this repo helps both you and me. If you find some mistakes, other novel or interesting works related to representation learning, please don't hesitate to issue or pull request.
- 1. Related Metrics
- 2. Related Survey
- 3. General Representation Learning (self-supervised, contrastive learning, ...)
- 4. Disentangled Representation
- 5. Disassembling Object Representation
- 6. VAE-based Method
- 7. GAN-based Method
- 8. Flow-based Method
- 9. Image Translation
- 10. Deep Clustering
- 11. Resources
- Evaluating the Disentanglement of Deep Generative Models through Manifold Topology, arXiv2020
- Theory and Evaluation Metrics for Learning Disentangled Representations, arXiv2019
- A framework for the quantitative evaluation of disentangled representations, ICLR2018
- Representation learning: A review and new perspectives, PAMI2013, Yoshua Bengio
- Recent Advances in Autoencoder-Based Representation Learning, arXiv2018
include self-supervised, contrastive learning, ...
-
A Large-Scale Study on Unsupervised Spatiotemporal Representation Learning, CVPR2021
codes-soon -
An Empirical Study of Training Self-Supervised Visual Transformers, arXiv2021, MoCov3[transformer]
-
Barlow Twins-Self-Supervised Learning via Redundancy Reduction, arXiv2021, Twins
codes-pytorch
-
Exploring Simple Siamese Representation Learning, arXiv2020, SimSiam
codes-pytorch -
Bootstrap your own latent-A new approach to self-supervised Learning, arXiv2020, BYOL
codes-pytorch -
Parametric Instance Classification for Unsupervised Visual Feature Learning, arXiv2020, PIC
-
A Universal Representation Transformer Layer for Few-Shot Image Classification, arXiv2020
codes-tensorflow -
Self-supervised Learning Generative or Contrastive, arXiv2020
-
Gradients as Features for Deep Representation Learning, ICLR2020
codes-pytorch -
Big Self-Supervised Models are Strong Semi-Supervised Learners, arXiv2020, SimCLR-v2
codes-tensorflow -
A simple framework for contrastive learning of visual representations, arXiv2020, SimCLR-v1
codes-pytorch -
Improved Baselines with Momentum Contrastive Learning, arXiv2020, MoCo-v2
codes-pytorch-unofficial -
Momentum Contrast for Unsupervised Visual Representation Learning, CVPR2020, MoCo-v1
codes-pytorch -
Self-supervised learning of pretext-invariant representations, CVPR2020, PIRL
codes-pytorch-unofficial -
Prototypical Contrastive Learning of Unsupervised Representations, arXiv2020
-
Supervised Contrastive Learning, arXiv2020
codes-pytorch -
Contrastive representation distillation. ICLR2020
codes-pytorch
-
Learning deep representations by mutual information estimation and maximization, ICLR2019
codes-pytorch -
Revisiting self-supervised visual representation learning, CVPR2019
-
Representation learning with contrastive predictive coding, arXiv2018
codes-keras -
Unsupervised feature learning via non-parametric instance discrimination, CVPR2018
codes-pytorch -
Contrastive learning of emoji-based representations for resource-poor languages
- ICAM: Interpretable Classification via Disentangled Representations and Feature Attribution Mapping, arXiv2020
codes
-
Challenging Common Assumptions in the Unsupervised Learning of Disentangled Representations, ICML2019
-
Emerging disentanglement in auto-encoder based unsupervised image content transfer, ICLR2019
codes-pytorch -
Are Disentangled Representations Helpful for Abstract Visual Reasoning?, NeurIPS2019
-
Hyperprior induced unsupervised disentanglement of latent representations, AAAI2019
codes-tensorflow
-
Learning disentangled joint continuous and discrete representations, NeurIPS2018
-
Unsupervised representation learning by predicting image rotations, ICLR2018
codes-pytorch -
A Two-Step Disentanglement Method, CVPR2018
codes-keras -
Disentangling by factorising, ICML2018
codes-pytorch -
Isolating sources of disentanglement in variational autoencoders, NeurIPS2018
codes-pytorch -
Life-long disentangled representation learning with cross-domain latent homologies, NeurIPS2018
-
A Spectral Regularizer for Unsupervised Disentanglement, arXiv2018
-
Visual object networks: Image generation with disentangled 3D representations, NeurIPS2018
-
Understanding disentangling in beta-VAE, arXiv2018
-
Disentangling the independently controllable factors of variation by interacting with the world, arXiv2018
-
Dual swap disentangling, NeurIPS2018
-
One-sample Guided Object Representation Disassembling, Neurips2020
codes-tensorflow -
Disassembling Object Representations without Labels, arXiv2020
-
Learning to Manipulate Individual Objects in an Image, CVPR2020
-
Monet: Unsupervised scene decomposition and representation, arXiv2019
codes-pytorch -
Multi-object representation learning with iterative variational inference, ICML2019
codes-pytorch -
Object Discovery with a Copy-Pasting GAN, arXiv2019
codes-tensorflow -
GENESIS: Generative Scene Inference and Sampling with Object-Centric Latent Representations, ICLR2019
codes-pytorch -
Stacked capsule autoencoders, NeurIPS2019
codes-pytorch -
LAVAE: Disentangling Location and Appearance, arXiv2019
-
Unsupervised object segmentation by redrawing, NeurIPS2019
codes-pytorch
- Relational neural expectation maximization: Unsupervised discovery of objects and their interactions, ICLR2018
codes-tensorflow
- Spatial broadcast decoder: A simple architecture for learning disentangled representations in vaes, arXiv2019
- beta-VAE Learning Basic Visual Concepts with a Constrained Variational Framework, ICLR2017
codes-pytorch
-
PixelGAN autoencoders, NeurIPS2017
-
Adversarial feature learning, ICLR2017
codes-theano
- Infogan: Interpretable representation learning by information maximizing generative adversarial nets, NeurIPS2016
codes-tensorflow
- Unsupervised representation learning with deep convolutional generative adversarial networks, arXiv2015
codes-tensorflow
-
High Resolution Face Age Editing, arXiv2020
codes-pytorch -
Rethinking the Truly Unsupervised Image-to-Image Translation, arXiv2020
codes-pytorch
-
Diverse Image-to-Image Translation via Disentangled Representations, ECCV2018
codes-tensorflow -
Image-to-image translation for cross-domain disentanglement, NeurIPS2018
codes-tensorflow
- Invariant information clustering for unsupervised image classification and segmentation, ICCV2019
codes-pytorch - Deep Spectral Clustering using Dual Autoencoder Network, CVPR2019
codes-tensorflow
-
Deep Clustering for Unsupervised Learning of Visual Features, ECCV2018
codes-pytorch -
Clustering with Deep Learning: Taxonomy and New Methods, arXiv2018
codes-theano
- disentangled-representation-papers, update 13 months ago
- Self-supervised learning and computer vision, 13 Jan 2020
- Self-Supervised Representation Learning, 10 Nov 2019