Liviu-Daniel's repositories
Dense_OpticalFlow_and_CNN_based_Motion_Segmentation_and_Object_Recognition
Farneback flow algorithm to segment motion and classify the ROI using CNN - Caffe model
caffe-recon-dec
Caffe for "Augmenting Supervised Neural Networks with Unsupervised Objectives for Large-scale Image Classification"
coursera-ML
use numpy and tensorflow to implement these basic ML model and learning algorithm
denseFlow_gpu
using GPU to extract videos' dense optical flow fields
deepmask
Torch implementation of DeepMask and SharpMask
labeld
LabelD is a quick and easy-to-use image annotation tool, built for academics, data scientists, and software engineers to enable single track or distributed image tagging. LabelD supports both localized, in-image (multi-)tagging, as well as image categorization.
Scripts
Collection of scripts, papers and notes
SqueezeNet
SqueezeNet: AlexNet-level accuracy with 50x fewer parameters
thresholding_methods
Thresholding methods for Multi Label Classification
dgd_person_reid
Domain Guided Dropout for Person Re-identification
caffe-video_triplet
Unsupervised Learning using Videos (ICCV 2015)
T-CNN
ImageNet 2015 Object Detection from Video (VID)
dtf_fisher
Improved Dense Trajectory features encoded by Fisher Vectors
faster_rcnn
Faster R-CNN
docker-fedora-opencv
Dockerfile for OpenCV build on Fedora
Recurrent-Convolutional-Video-ReID
Implementation of our paper - Recurrent Convolutional Network for Video-based Person Re-Identification
docker-quick-ref
Docker: Printable Quick Reference
image_captioning
Image Captioning with Deep Bidirectional LSTMs
kaggle-chars74k
Classification of characters from Google Street View images
Tools-for-Caffe
Various tools useful for playing with Caffe written in Python
clicker
Store click positions for a dataset of images (machine learning labeling)
KinectHandTracking
A simple hand tracking and facial recognition program using Kinect 2 SDK, with the ability to interact with desktop and other third party program.
caffe-augmentation
:zap: Caffe real-time data augmentation on-the-fly!!
New-C3D-Caffe
This is a reimplementation of 3D CNN (http://vlg.cs.dartmouth.edu/c3d/). It is compatitable with Caffe 2016. The Caffe is forked from Caffe commit c1126aadb6eab5229203bfba9495bc2b06b4ae89.