Computer VisionArtificial IntelligenceMediaPipeOpenCVStreamlitFitness Tech
AI Real-Time Gym Coach
Real-time AI fitness coach that uses computer vision and pose estimation to track exercises, count reps, analyze form, and provide live coaching feedback.

Real-time
Pose Analysis
Live
Workout Tracking
Enabled
AI Coaching
Problem statement
Many people train without a personal trainer and may struggle to maintain correct exercise form, track repetitions accurately, or recognize movement mistakes during workouts.
Solution
Built a real-time computer-vision fitness coach using MediaPipe pose estimation, OpenCV, WebRTC, and exercise-specific detection logic. The app analyzes body landmarks and movement patterns to track reps and sets, evaluate form-related metrics, and provide AI-powered coaching and voice feedback.
Key features
- Real-time webcam pose estimation
- Automatic repetition counting
- Set and workout tracking
- Exercise-specific form analysis
- Real-time form feedback
- AI-powered coaching with Groq
- Voice coaching and text-to-speech feedback
- Multiple exercise detectors
- Workout history and progress tracking
- Interactive Streamlit interface
Tech stack
PythonMediaPipeOpenCVStreamlitWebRTCGroqgTTS
Screenshots






Future improvements
- Add more exercise detectors
- Improve form classification
- Personalized workout plans
- Advanced workout analytics
- Progress visualization
- Calorie estimation
- Multilingual AI coaching
- Mobile app version