This project is a multi-modal AI system for identifying badminton talent using:
- Text-based personal data
- Pose detection from videos
- Match outcome prediction
- Talent recommendation based on physical & personal attributes
- Pose analysis using MediaPipe
- Lightweight ML models
- Simple and clean UI
- Python
- Scikit-learn
- MediaPipe
- FastAPI
- Streamlit
- Text-based recommendation
- Pose-based motion analysis
- Match outcome prediction
Due to lack of public Iranian badminton datasets, part of the data is synthetic or user-generated.