AI-Powered Video Analysis for 50–100 Soccer Training Exercises

This project aims to develop an AI-powered video analysis system for a mobile soccer training application. The system will analyze user-uploaded video recordings of soccer drills to track the ball and player movements. Key objectives include detecting repetitions, sequences, rule compliance for exercises like juggling and kicking drills. The analysis will output metrics such as repetition count, time taken, foot usage to provide automated feedback. Developers are sought to build a prototype analyzing 5 drills, scaling to 100, along with modular code, an API and accuracy benchmarks integrating with a React Native/Flutter app concept. Strong skills in computer vision, pose estimation, objection detection with YOLOv8 and mediapipe are required. The deliverables will be milestone-driven with an open budget discussed.