PoseTrack AI 🏋️‍♂️

Real-Time AI Fitness Coach Using Computer Vision & Pose Estimation

PoseTrack AI is a full-stack AI-powered fitness tracking platform that uses computer vision and pose estimation to analyze body movements in real time. The system tracks exercises, counts repetitions automatically, provides form feedback, calculates calories burned, and maintains workout history through an interactive web dashboard.

Built using Flask, MediaPipe, OpenCV, and modern web technologies, PoseTrack AI acts as a virtual fitness assistant capable of helping users monitor and improve workout performance directly through their webcam.

🎥 Demo

▶️ Click the image to watch the full demonstration

🚀 Key Features

🤖 AI-Powered Pose Detection

Real-time body landmark detection using MediaPipe
Accurate joint angle calculation
Live pose tracking through webcam

💪 Intelligent Exercise Tracking

Supports automatic repetition counting for:

Bicep Curl
Squat
Push Up
Shoulder Press
Lunge
Deadlift
Leg Raise
Lateral Raise

📊 Workout Analytics

Daily workout statistics
Exercise history tracking
Calories burned estimation
Performance monitoring

👤 User Management

Secure authentication system
Personal workout records
Individual fitness progress tracking

📁 Data Export

Export workout history to CSV
Download performance reports

🏗️ System Architecture

User Webcam
      │
      ▼
MediaPipe Pose Detection
      │
      ▼
Body Landmark Extraction
      │
      ▼
Joint Angle Calculation
      │
      ▼
Exercise Recognition Logic
      │
      ▼
Repetition Counter
      │
      ▼
Calorie Estimation Engine
      │
      ▼
Workout History Storage
      │
      ▼
Analytics Dashboard

🛠️ Technology Stack

Backend

Python
Flask
Gunicorn

Artificial Intelligence & Computer Vision

MediaPipe
OpenCV
NumPy

Frontend

HTML5
CSS3
JavaScript
Bootstrap

Database & Storage

SQLite
CSV Export Support

Supported Environment

✅ Full functionality available on Localhost

Real-time camera access
Live pose detection
Exercise tracking
Rep counting
Form analysis

To Experience the Complete Application

Clone the repository and run it locally:

git clone https://github.com/pn-dev-in/AI-Powered-Fitness-Tracking-System.git
cd AI-Powered-Fitness-Tracking-System
pip install -r requirements.txt
python app.py

Then open:

http://127.0.0.1:5000

This provides full access to all AI-powered exercise tracking capabilities.

📂 Project Structure

PoseTrack-AI/

├── controllers/
├── models/
├── services/
├── static/
├── templates/
├── tests/
├── utils/
├── user_data/
│
├── app.py
├── auth_utils.py
├── config.py
├── setup_db.py
│
├── Dockerfile
├── docker-compose.yml
├── render.yaml
├── requirements.txt
└── README.md

🎯 Supported Exercises

Exercise	Calories/Rep	Difficulty
Bicep Curl	0.5	Beginner
Squat	1.2	Intermediate
Push Up	1.0	Intermediate
Shoulder Press	0.8	Beginner
Lunge	1.0	Intermediate
Deadlift	1.5	Advanced
Leg Raise	0.7	Beginner
Lateral Raise	0.4	Beginner

📸 Application Screenshots

Signup Page

Signin Page

Dashboard

Live Exercise Tracking

AI Coach

Analytics Dashboard

⚡ Local Installation

Clone the repository:

git clone https://github.com/pn-dev-in/AI-Powered-Fitness-Tracking-System.git

cd AI-Powered-Fitness-Tracking-System

Create virtual environment:

python -m venv venv

Activate environment:

Windows

venv\Scripts\activate

Linux / Mac

source venv/bin/activate

Install dependencies:

pip install -r requirements.txt

Initialize database:

python setup_db.py

Run application:

python app.py

Open:

http://127.0.0.1:5000

🐳 Docker Setup

Build container:

docker build -t posetrack-ai .

Run container:

docker run -p 5000:5000 posetrack-ai

Using Docker Compose:

docker-compose up --build

🧪 Testing

Run project tests:

pytest

📈 Future Enhancements

AI posture correction feedback
Exercise auto-classification using ML
Personalized workout recommendations
Fitness goal tracking
Mobile application
Voice-enabled AI fitness coach
Wearable device integration
Advanced analytics and reporting

🎓 Skills Demonstrated

This project showcases practical experience in:

✅ Computer Vision

✅ Pose Estimation

✅ Artificial Intelligence

✅ Flask Web Development

✅ RESTful Application Design

✅ Authentication Systems

✅ Data Analytics

✅ Docker Containerization

✅ Cloud Deployment

✅ Software Architecture

📊 Project Highlights

Real-time AI fitness monitoring
8 supported exercise types
Automated repetition counting
Live calorie estimation
User authentication and session management
Cloud-hosted production deployment
Docker-ready infrastructure
Exportable workout data

👨‍💻 Author

Pravesh Nandanwar

GitHub: https://github.com/pn-dev-in

LinkedIn: www.linkedin.com/in/pravesh-nandanwar

⭐ Support

If you found this project useful, consider giving it a star on GitHub.

Contributions, suggestions, and feedback are always welcome.

Name		Name	Last commit message	Last commit date
assets		assets
controllers		controllers
flask_session		flask_session
models		models
services		services
static		static
templates		templates
tests		tests
user_data		user_data
utils		utils
.env.example		.env.example
.gitignore		.gitignore
.renderignore		.renderignore
Dockerfile		Dockerfile
README.md		README.md
app.py		app.py
auth_utils.py		auth_utils.py
check_users.py		check_users.py
config.py		config.py
docker-compose.yml		docker-compose.yml
gunicorn.conf.py		gunicorn.conf.py
render.yaml		render.yaml
requirements.txt		requirements.txt
runtime.txt		runtime.txt
setup_db.py		setup_db.py

Latest commit

History

Folders and files

Repository files navigation