This is a real-time AI project that can detect and count bicep curls from a live video feed using computer vision. No need for wearables or fitness devices — just your camera and the power of AI.
Currently, the system supports bicep curls only, but is designed to be easily extended to support other exercises in future updates.
- Uses your webcam or any connected camera to detect your body movements.
- Tracks the angle of your arm to determine when a full curl is completed.
- Counts reps live on-screen and can be adjusted for different difficulty levels.
- Python – Base programming language
- OpenCV – Video capture and real-time processing
- MediaPipe – Pose detection and landmark tracking
- CVZone – Simplified overlays and UI
- TensorFlow – Trained custom model (future use for multi-exercise support)
- ✅ Real-time bicep curl detection
- ✅ Accurate rep counting based on elbow angles
- ✅ Visual UI overlays (count, curl stage, etc.)
- ✅ Works with any USB webcam or built-in laptop camera
- 🔜 More exercises and real-time feedback features coming soon!
- Clone the repository:
git clone https://github.com/yourusername/ai-bicep-curl-counter.git cd ai-bicep-curl-counter