We built an open-source clothing recognizer that processes photos in your browser

We're building Vestes Archive, and have released Pasma: a small clothing recognizer for wardrobe and resale apps. Choose a photo of one garment and get three likely garment types, a category and visual material suggestions, then correct anything it gets wrong. The demo runs on your computer in the browser. No account or API key is required, and your photo isn't uploaded. First use downloads the model weights and runtime. It's a v1 candidate rather than a complete wardrobe app. It doesn't identify brands or certify fabric composition, and multiple garments in one photo can confuse it. Developers can use the open weights and ONNX/CoreML exports under Apache-2.0. Try it: huggingface.co/spaces/Prikol5ist/pasma Code and 25-second demo video: github.com/prikol5ist-art/Pasma Weights: huggingface.co/Prikol5ist/pasma Would this be useful while cataloguing clothes, or would another step in that workflow save you more time?