Jim Clyde Monge
Built at Meta Global AI Developer Hackathon · Oct 26, 2026 · Remote

ReScene turns product photos into interactive 3D shopping experiences. Online shopping usually means scrolling through photos, opening separate product pages to view it in different angles, and trying to imagine how it would look when placed together with other products. ReScene lets shoppers turn products around, preview colors, arrange items together, and ask an AI stylist for combinations within their budget. Businesses can start with an existing photo. ReScene identifies and segments its objects, reconstructs individual 3D models, and assembles an editable environment. Users can adjust placement, lighting, colors, and textures, then attach product names, descriptions, prices, and shopping links. Those assets and details become a reusable catalog. Alternatively, users can select product photos and a background, describe a setting, and generate a new scene with Muse Image before reconstructing it in 3D. The working shopping demo includes three experiences: 1. Explore a Product: browse product photography, switch to interactive 3D, and preview colors. 2. Build Your Space: arrange products, replace items, change the environment, and request affordable combinations through text or voice. 3. Shop the Scene: explore a complete product setup in 3D, or select products directly in the original photo using saved segmentation masks. I built ReScene from scratch during the hackathon. It demonstrates how retailers could use the same catalog across product pages, interactive rooms, and shoppable images.