CS 838 Learning Based Methods in Computer Vision

Violence Detection in Surveillance from Drones

This repository contains the deployment of several modules used for the project ErgoCub

Designing a 3D CNN architecture for action recognition in video

Action temporal label tool, for pose skeleton based CNN action recognition etc.

Master Internship project

Action recognition using LSTM

基于跨尺度图对比学习的人体骨架动作识别方法(Cross-Scale Graph Contrastive Learning framework for Skeleton-based action Representation, CrosScale-SGCLR)

Navigate Youtube passively, skip ads and non-interesting fragments of interesting videos - per topic or from watch list

PyTorch implementation of a Relational Attention-Based Model for Group Activity Recognition — inspired by "A Hierarchical Deep Temporal Model for Group Activity Recognition (arXiv:1511.06040v2)"

Code for VideoSSL: Semi-Supervised Learning for Video Classification.

AVT Dissertation Research, Architecture used for AVT experimentation, mainly with backbone model and attention heads.

Spatial Temporal Graph Convolutional Networks (ST-GCN) for Skeleton-Based Action Recognition in PyTorch

Code for the ICPR2020 Oral paper "Subspace Clustering for Action Recognition with Covariance Representations and Temporal Pruning"

Use 45 Feature to recognition the class of the action.

Final project for Nile university CIT-651 Machine learning class - Fall 19

Dataset repository of "MetaVD: A Meta Video Dataset for enhancing human action recognition datasets."

Implemented a CNN-LSTM Action Recognizer for dynamic motion analysis, integrating convolutional and recurrent neural networks to efficiently recognize and classify actions in video data of UCF101 dataset.