Transform your fitness journey with the power of AI. FitHub is a modern, full-stack web application designed to provide personalized workout guidance, real-time pose correction, and intelligent nutrition planning, all tailored to your unique body type and goals.
Key Features • AI Models • Tech Stack • Getting Started • Screenshots
- Three Core AI Models: Specialized modules for body type analysis, diet planning, and real-time workout training.
- User Authentication: Secure user registration and login system to manage your personal dashboard and track progress.
- Interactive Dashboard: A central hub to view your stats, access AI models, and get workout recommendations.
- Fully Responsive Design: A sleek, modern, and dark-themed UI that looks great on any device, from mobile phones to desktops.
FitHub's intelligence is powered by a suite of specialized machine learning models running on a Python backend. Each model is designed to tackle a specific aspect of your fitness journey.
- Purpose: To classify a user's physique into one of the three main somatotypes: Ectomorph, Mesomorph, or Endomorph.
- How it Works: The model is a Convolutional Neural Network (CNN) trained on a large dataset of body images. It analyzes a user-submitted photo to identify key physical features and predicts the most likely body type, providing a confidence score for its classification.
- Technology: Python, TensorFlow/Keras, OpenCV for image preprocessing.
- Purpose: To generate personalized meal plans based on user-provided data.
- How it Works: This model uses a combination of rule-based algorithms and machine learning to create a diet plan. It considers the user's body type, BMI, fitness goals (e.g., weight loss, muscle gain), and dietary preferences to recommend meals that meet specific caloric and macronutrient targets.
- Technology: Python, Scikit-learn, Pandas for data manipulation.
- Purpose: To provide real-time feedback on exercise form during a workout session.
- How it Works: This feature utilizes a real-time pose estimation model. It processes the user's webcam feed to map out 33 key body landmarks (joints, limbs, etc.). By analyzing the angles and positions of these landmarks, it can determine if an exercise is being performed correctly and provide instant corrective feedback.
- Technology: Python, OpenCV, MediaPipe for pose estimation.
This project is a full-stack application built with the MERN stack and other modern technologies.
| Category | Technology |
|---|---|
| Frontend | React, TypeScript, Vite, Tailwind CSS, Shadcn/UI |
| Backend | Node.js, Express.js |
| Database | MongoDB with Mongoose |
| AI / Machine Learning | Python, TensorFlow, OpenCV, MediaPipe |
| Auth | JSON Web Tokens (JWT), bcrypt.js for password hashing |
| Styling | Tailwind CSS, Lucide React for icons |
Follow these instructions to get a copy of the project up and running on your local machine for development and testing purposes.
- Node.js (v18.x or later)
- npm (usually comes with Node.js)
- MongoDB installed and running locally.
- Python (v3.8 or later) with relevant ML libraries.
-
Clone the repository:
git clone [https://github.com/Shiva2806/FitHub.git](https://github.com/Shiva2806/FitHub.git) cd FitHub -
Set up the Backend Server:
- Navigate to the
serverdirectory:cd server - Install the dependencies:
npm install
- Create a
.envfile in theserverdirectory and add your MongoDB connection string:MONGODB_URI=mongodb://localhost:27017/fithub JWT_SECRET=your_jwt_secret_key_here
- Start the backend server:
npm run dev
Your backend should now be running on
http://localhost:5000. - Navigate to the
-
Set up the Frontend Client:
- Open a new terminal and navigate to the
clientdirectory:cd client - Install the dependencies:
npm install
- Start the frontend development server:
npm run dev
Your frontend should now be running on
http://localhost:3000(or another port if 3000 is busy). - Open a new terminal and navigate to the
-
You're all set! Open your browser and navigate to the frontend URL to start using FitHub.
Homepage
AI Models Selection Page
User Dashboard