Classify any image, in-browser.

This is a general-purpose vision classifier, not a diagnostic model — we're showing the mechanics of an in-browser imaging pipeline (upload → preprocess → inference → ranked output), the same shape as the modality-specific models we build for imaging engagements.