This project is an AI-powered virtual dressing room that allows users to try on clothes in real time using a live camera feed. Built using Django and React, the platform leverages OpenCV and Mediapipe for real-time garment overlay, giving users an interactive virtual try-on experience.
✅ Real-time garment overlay on live video feed
✅ Accurate garment fitting with shoulder width adjustment
✅ Hand gesture-based garment selection
✅ Horizontal gallery for garment selection
✅ Scalable architecture using Django + React
✅ Fast communication using WebSockets
-
Frontend (React):
- Captures video frames from the user's camera using Mediapipe.
- Sends frames to the backend via WebSockets.
-
Backend (Django):
- Processes incoming frames using OpenCV.
- Overlays the selected garment onto the video frame.
- Adjusts garment size based on dynamic
tshirt_shoulder_widthfor a perfect fit.
-
Garment Selection:
- Left and right windows display garment previews.
- Middle window shows the live try-on with the overlay.
- Bottom row gallery allows users to scroll and select garments.
- Hand gesture detection allows garment selection.
| Technology | Purpose |
|---|---|
| Django | Backend framework |
| React (Vite) | Frontend framework |
| OpenCV | Computer vision for garment overlay |
| Mediapipe | Hand gesture recognition |
| WebSockets | Real-time communication |
| PostgreSQL | Database for storing garment images (in bytea format) |
git clone https://github.com/your-username/virtual-dressing-room.git
cd virtual-dressing-room