SL2T

A sign-language-to-text translation model: video of signing (processed as pose landmark coordinates by on-device MediaPipe Holistic) in, spoken-language text out. Trained on over 100,000 hours across 50+ sign languages; scores 70 BLEURT zero-shot on FLEURS-ASL.

Работает в Gboard (диктовка жестами) и Live Transcribe, старт на Pixel 11 с американского жестового на английский. Анонс 12.08.2026.

Understands

video, text

Produces

text

Contents verified with the vendor: 2026-08-18

Included in subscriptions

No catalogue entry marks this model as part of a plan yet. It will appear here once the plan contents are confirmed with the vendor.