# RightOnTOp

- **Event:** [Meta Global AI Developer Hackathon](https://cerebralvalley.ai/e/meta-ai-developer-hackathon)
- **When:** Oct 26, 12:00 PM – Nov 4, 9:00 PM (EST)
- **Where:** Online
- **Team:** [Ajay L](https://cerebralvalley.ai/u/Ajay55m)
- **GitHub:** https://github.com/ajay55m/VirtualTrialRoom
- **Demo video:** https://drive.google.com/file/d/1FD3nsPjA-Rk7ltspcMMuG6c5ufbJrekb/view?usp=drive_link
- **Gallery:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery
- **Page:** https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2397

VirtualFit AI Studio is a Universal Virtual Trial Room that lets any user try any outfit on their own photo in seconds - Shirts, T-Shirts, Pants, Jeans, Dresses, and Full-Body Sets - without visiting a physical store.

Problem: Online fashion has a 30-40% return rate because customers cannot visualize how an outfit will look and fit on their own body. Existing virtual try-on tools only work for simple t-shirts, fail on full sets and traditional wear, and distort the face, hair, and body shape.

Solution: We built a Universal 3-Stage AI Engine:
1. Scope 1 - Analyzer: Detects garment type, category (Upper-Body / Lower-Body / Full-Body), color and style from any seller product photo.
2. Scope 2 - Standardizer: Cleans any product photo, removes background/model, and creates a try-on ready garment. For sets, it splits Top and Bottom.
3. Scope 3 - Universal Try-On: Uses an innovative 2-Pass System for full sets. Pass-1 dresses the Top (Jacket/Kurta), Pass-2 dresses the Bottom (Trouser/Pajama) on the Pass-1 result. This preserves the user's face, hair, body, pose and shoes 100%.

Key Features: One photo try-on, support for Men/Women/Kids, Try Another Outfit carousel to try 12+ outfits on the same model without re-uploading, and a Before/After drag slider.

Impact: It brings the physical trial room online, increases buying confidence, and helps e-commerce sellers and D2C brands reduce returns and increase conversions.

## More from Meta Global AI Developer Hackathon

- [Croissants enjoyer](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2394)
- [CanisMajor](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2395)
- [Muse RAISE](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2396)
- [Relay](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2398)
- [AuditCanvas](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2399)
- [self](https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2400)

---

Markdown version of https://cerebralvalley.ai/e/meta-ai-developer-hackathon/hackathon/gallery/2397. Site index for agents: https://cerebralvalley.ai/llms.txt · full text: https://cerebralvalley.ai/llms-full.txt