From photos to outfits
Codex crops each garment from a photo. GPT Image reconstructs the item on a solid-color background, which can then be removed to create a clean cutout. Those pieces become a wardrobe to browse, combine, and render as outfits. My explanation in the replies ↗
