wtf is outfitanyone?

OutfitAnyone is an AI research project from Alibaba Group's Institute for Intelligent Computing that lets you see what a piece of clothing would look like on a person without physically trying it on. You provide a clothing image, and the system composites it onto a person using a generative AI model, producing a realistic-looking result. The current public demo, available on Hugging Face and on ModelScope for users in China, only accepts clothing images as input. The people shown in the demo outputs are pre-generated AI figures rather than real uploaded photos. The README notes this restriction is in place to prevent misuse of real personal photos. The project is research-level work published as an academic paper on arXiv. The repository accompanies that paper and provides access to a demo rather than a full self-hosted codebase for running locally. There is also a companion demo combining OutfitAnyone with a separate project called AnimateAnyone, which adds motion to the try-on results. The README is short and does not include installation instructions or model weights. It primarily links to the project page, the paper, and the hosted demos. If you want to try the system, the Hugging Face Space is the most accessible entry point.