StyleMirror AI

See yourself in every outfit before you buy with AI-powered virtual try-on and a personal Gemini fashion stylist.

Comment

StyleMirror AI

StyleMirror AI is an AI-powered virtual fashion try-on experience that helps shoppers see how clothing could look on them before making a purchase.

Inspiration

Online clothing shopping can be difficult because customers cannot physically try on clothes before buying. We wanted to create a simple experience where users can upload their photo, select an outfit, and instantly see themselves wearing it.

What we built

StyleMirror AI combines virtual try-on and AI-powered fashion assistance into one workflow:

  1. Upload Photo — Users upload a front-facing photo.
  2. Select Clothing — Users browse a catalog of clothing items and select an outfit.
  3. Virtual Try-On — The selected clothing image and user's photo are sent to the YouCam Apparel Virtual Try-On API.
  4. AI Fashion Stylist — Google Gemini analyzes the selected outfit and provides personalized styling advice.
  5. Shop Now — Users can view the selected product information and continue to shopping.

The product catalog is served through our backend API, while the frontend provides a responsive interactive experience.

How we built it

The frontend was built with React and Vite, with Lucide icons and custom responsive styling.

The backend was built with Node.js and Express. It provides product catalog APIs, the virtual try-on endpoint, and the AI stylist endpoint.

We integrated the YouCam Apparel Virtual Try-On API to generate the virtual try-on experience and Google Gemini to power the AI fashion stylist.

The application is deployed using Vercel, with the frontend and backend connected through environment variables and API endpoints.

Challenges

One of the biggest challenges was connecting the frontend, backend, third-party APIs, and production deployment correctly. We also had to handle CORS configuration, environment variables, API responses, and external clothing image URLs.

During development, we tested the application locally and then configured the production environment so that the same workflow works from the deployed application.

What we learned

This project helped us understand how AI APIs can be combined to create a complete user-facing product rather than a standalone AI feature. We learned about virtual try-on workflows, API integration, frontend/backend communication, CORS, deployment, and building an AI-powered conversational styling experience.

Future plans

In the future, we want to add real e-commerce product integrations, personalized recommendations based on user preferences, body and size recommendations, saved outfits, shopping history, and support for more clothing categories.

StyleMirror AI aims to make online fashion shopping more confident, interactive, and personalized.

Built With

Share this project:

Updates