This project demonstrates an AI-based football video analysis system built using supervision, roboflow, and Hugging Face vision models. The system processes match footage to detect, track, and analyze player movement, ball position, and tactical patterns.
It combines computer vision, deep learning, and geometric transformations to provide a detailed view of player behavior and match dynamics.