SkalskiP/sports
Jupyter notebooks demonstrating football player detection, tracking, and 3D pose estimation using YOLOv5, YOLOv7, and ByteTrack with PyTorch.

Velocity · 7d
+0.4
★ / day
Trend
→steady
star history
This repository contains Jupyter notebook tutorials showcasing computer vision applications in sports analytics. It demonstrates player tracking using YOLOv5 combined with ByteTrack for multi-object tracking, and 3D pose estimation using YOLOv7 for analyzing player positioning. The experiments are applied to football/soccer footage for applications like VAR offside analysis.