skeleton-based

[TIFS 2019] Skeleton-based Gait Recognition via Robust Frame-level Matching (RFM)

Official Pytorch Implementation of our paper: GAF-Net: Video-Based Person Re-Identification via Appearance and Gait Recognitions

A deep learning-based system for detecting weight training form errors using 3D-CNN enhanced with a novel biomechanical attention mechanism. This project includes a full pipeline from video-based pose estimation (MediaPipe) to skeleton-based action recognition and future error classification.

ecommerce

Real-time skeleton-based action recognition with from-scratch HRNet-like pose estimation and two-stream ST-GCN, targeting NTU RGB+D 120