LayerLook is an AI-powered fashion storytelling tool that helps creators transform outfit photos, accessories, fit checks, and fashion references into polished editorial-style collages optimized for social media. Built during the UPSCALE Conference Hackathon.
Built by Team Laputa (Peggy Shen & Bernice Sun) for the UPSCALE Conference Hackathon — June 2026.
🎥 Final V2 Demo: https://youtu.be/yngMoPg11wM
🎥 V1 Demo: https://drive.google.com/file/d/1Qis1on8kP9NqD5IxHLb1cjAEuQYwNKQO/view?usp=sharing
Fashion creators spend a surprising amount of time manually creating outfit collages and story posts.
A typical workflow involves collecting outfit photos, removing backgrounds, cutting out accessories, arranging layouts, resizing assets, and experimenting with aesthetics until everything feels ready to post.
While collecting inspiration is fun, turning that inspiration into a polished post is still largely manual.
LayerLook helps bridge that gap.
LayerLook transforms fashion assets into social-ready visual stories.
Users can:
- Upload outfit photos and fashion references
- Choose a creative direction
- Generate a polished fashion composition
The same outfit can be interpreted through multiple visual styles, helping creators quickly explore different aesthetics.
Transform outfit photos into curated visual compositions.
Automatically isolate fashion subjects using Google Gemini-powered image processing.
Generate stylistically distinct outputs:
- Editorial magazine layouts
- Scrapbook collages
- Visual journals
- Modernist compositions
- Social-first fashion stories
Designed for content formats such as:
- Instagram Stories
- Outfit recaps
- Fashion moodboards
- Pinterest-style collages
Used for:
- rapid prototyping
- frontend development
- deployment
- full-stack workflow
Used for:
- image processing
- background removal
- transparent cutout generation
Potential applications:
- fashion image understanding
- garment recognition, accessory detection, image segmentation
- style analysis
Potential applications:
- image enhancement, upscaling
- texture refinement
- editorial-quality output polishing
UPSCALE AI Conference Hackathon San Francisco, California June 2026
Made by Bernice Sun and Peggy Shen.
"Spend less time cutting out images. Spend more time creating."