📌 Project Overview
Meta-Closet is an AI-powered virtual outfit try-on system that allows users to visualize how clothes would fit them without physically wearing them. It is a simple extension feature which works on various different e-commerce platform such as ammazon, meesho ect.
Feaaturing in the improvement of customer service and their engagement while shopping. This project involves ✨ Key Features
✅ Real-Time Virtual Try-On
- Uses AI-powered body scanning and AR to overlay clothes on the user.
- Works with both live camera feed & pre-captured images.
✅ Body Measurement & Fit Estimation
- Detects height, weight, and body shape from a simple scan.
- Suggests best-fitting sizes automatically.
✅ E-Commerce & Brand Integration
- Can be integrated with fashion websites (Amazon, Myntra, Ajio, etc.).
- Users can buy clothes directly after trying them.
🛠️ Tech Stack
🔹 Frontend (User Interface & AR Integration)
- React.js + Vite – Web app UI
🔹 Backend (AI Processing & User Data Management)
- Python (FastAPI / Flask / Django) – Backend API
🔹 AI & Computer Vision (Core Try-On Feature)
- OpenPose – Body pose detection
- Hugging face Cloth segmentation– AI-powered outfit segmentation
- VITON-HD – Realistic outfit overlay
- PyTorch – Model training
🔹 Cloud Services & Deployment
💡 Where Can Meta-Closet Be Used?
✅ E-Commerce Platforms – Virtual try-on for fashion brands. ✅ Shopping Mall Kiosks – In-store virtual trial rooms. ✅ Fashion Design & Tailoring – Custom clothing fitting. ✅ Social Media & Influencers – AI-driven styling & recommendations.
🚀 Future Scope & Enhancements
🔹 Virtual Wardrobe – Users can save and manage their outfits. 🔹 Smart Mirror Integration – A touchless in-store try-on experience.
📢 Conclusion
Meta-Closet is a revolutionary fashion-tech solution that bridges the gap between online shopping and real-world outfit trials. It makes shopping easier, smarter, and more interactive while reducing returns and improving user satisfaction.
HOW TO INSTALL:
-
Download and configure cmake
-
Install the neccessary requirements.txt using pip pip install -r requirements.txt
-
Download the neccessary model weights from the Model weights