A Personalized AI-Powered Fitness Assistant
With so many unhealthy choices in daily life, we wanted to create a dedicated fitness platform that provides personalized workouts, making it easier for people to stay healthy and active.
IntelliFit helps users track their fitness progress and improve workout form through AI-driven pose estimation. Features include:
✅ User Accounts – Track fitness progress over time.
✅ Workout Selection – Choose from predefined workouts or create custom workouts to fit personal goals.
✅ AI-Powered Pose Estimation – Uses computer vision to overlay real-time posture corrections on the user's figure.
✅ Repetition Counter & Form Accuracy Score – Automatically detects and counts reps while assessing exercise form.
- React.js – A JavaScript framework for building dynamic web applications.
- Firebase – Used for user authentication and storing workout data.
- CSS (Styled Components / Tailwind CSS) – Enhances the UI with a modern, responsive design.
- Python – Powers the backend logic for AI-driven pose estimation.
- MediaPipe – A Google library for real-time pose estimation, enabling workout tracking and posture correction.
- Flask – Handles API communication between the frontend and backend.
- OpenCV – Used for processing real-time images and enhancing pose tracking.
- Custom AI Model Training – Each exercise is individually calibrated for accurate posture correction.
- Integrating Frontend & Backend – Handling image transmission between React and Python while ensuring real-time feedback.
- Calibrating the AI Model – Each exercise required separate fine-tuning for accurate pose estimation and rep counting.
- Successfully implemented AI-powered pose estimation to assist with real-time form correction.
- Built a fully functional fitness tracking platform with personalized workout creation.
- Overcame technical difficulties in frontend-backend communication to enable smooth user interactions.
- The complexity of real-time image processing and computer vision-based pose estimation.
- How challenging frontend-backend integration can be, especially when dealing with real-time data transmission.
- Fine-tuning AI models for different exercise types is more complex than expected but essential for accuracy.
- Enhancing Pose Estimation – Improve AI accuracy for better form correction and more reliable rep counting.
- Adding More Exercises – Expand the exercise library to cover more complex workout routines.
- Optimizing API Communication – Improve efficiency in sending workout images between the frontend and backend.
- Node.js and npm installed for frontend development.
- Python 3.x installed for backend development.
- Firebase Account for managing user authentication and workout data.
- Clone the repository:
git clone https://github.com/sanj6y/IntelliFit.git cd IntelliFit