A web application for real-time yoga pose detection and feedback using MediaPipe and WebSocket communication.

• Python backend with MediaPipe pose detection
• Real-time pose landmark detection and visualization
• Pose comparison and alignment feedback
• Calibration and session management
• Modern Next.js frontend with real-time feedback

1. Install Python dependencies:

   cd backend
   pip install -r requirements.txt

2. Test MediaPipe installation:

   python test_mediapipe.py

   This will open your webcam and verify MediaPipe is working correctly.

3. Start the WebSocket server:

   python app.py

   This will start the server at ws://127.0.0.1:5000.

1. Install Node.js dependencies:

   cd frontend
   npm install

2. Start the development server:

   npm run dev

   This will start the frontend at http://localhost:3000.

1. Open http://localhost:3000 in your browser
2. Navigate to the Workouts page
3. Allow webcam access when prompted
4. Calibrate each pose by clicking "Calibrate Pose"
5. Start a session by clicking "Start Session"
6. Follow the alignment feedback to improve your pose

• If the webcam doesn't appear, check browser permissions
• If landmarks aren't detected, ensure you're visible in the camera frame
• For backend issues, check the terminal for error messages
• For MediaPipe installation issues, run python test_mediapipe.py

• /backend - Python backend with MediaPipe and WebSocket server

  • app.py - Main WebSocket server with pose detection
  • test_mediapipe.py - Test script for MediaPipe functionality
  • calibration/ - Reference pose images

• /frontend - Next.js frontend application

  • app/workouts/page.tsx - Main page for yoga pose detection
  • src/hooks/useWebSocket.ts - WebSocket connection handling
  • public/poses/ - Reference pose images

1. The backend captures video from your webcam and processes it with MediaPipe
2. Real-time pose landmarks are extracted and sent via WebSocket
3. The frontend displays your webcam feed and overlays the landmarks
4. During calibration, your pose is saved as a reference
5. During a session, your current pose is compared to the calibrated references
6. Real-time feedback helps you adjust your alignment

For more detail on the workings of the application, check out these key files:

• backend/app.py - Main server implementation with MediaPipe integration
• frontend/app/workouts/page.tsx - Main frontend interface for pose detection
• frontend/src/hooks/useWebSocket.ts - WebSocket communication