Educational bidirectional communication between deaf and hearing people using American Sign Language (ASL).
- Type → Sign — Type English text; a skeleton signer animates ASL using the sign.mt pose pipeline (
pose-viewer). - Sign → Text — Sign supported vocabulary words on camera; words appear in a live transcript (browser-local MediaPipe).
- Next.js 15, React 19, TypeScript, Tailwind CSS
- pose-viewer — skeleton playback (same as sign.mt)
- MediaPipe Hands — in-browser word recognition
npm install --legacy-peer-deps
npm run devOpen http://localhost:3000.
Text-to-sign animation data is generated via the public sign.mt cloud function, as used by sign/translate.