Google DeepMind Launches SL2T: Real-Time Sign Language to Text on Pixel
Kyrannio · x · 2026-08-12
Google DeepMind has launched SL2T (Sign Language to Text), an AI model that directly translates sign language into text. Unlike traditional dictation tools that rely on audio, sign languages possess unique grammars and complex 3D movements.
Instead of using raw camera feeds or physical gloves, SL2T utilizes MediaPipe Holistic to track body landmark coordinates locally and translates these spatial points into text. The technology is now powering sign-to-text dictation on Pixel 11 in Gboard and Live Transcribe, starting with ASL. Users can now sign naturally to search, draft messages, or engage in live conversations.
Related event: DeepMind Launches SL2T Sign Language Model on Pixel 11(10 posts)→
More from Apps
- Gamma Hits $100M ARR via AEO: Podcast Dives into AI Search Growth Tactics — vasuman · 2026-08-13
- Google AI Overviews fail: local search citing low-quality self-promotional listicles — gaganghotra_ · 2026-08-13
- LangSmith launches BYOC on AWS, keeping enterprise agent data inside private VPCs — LangChain · 2026-08-13
- Google Search to Lower Follower Thresholds for Creator Profiles, Expand Globally — rmstein · 2026-08-13
- Dev Builds Interactive Solar Eclipse Simulation via Google AI Studio — AI_Andrew · 2026-08-13
- Developer Recreates Cinematic 'Odyssey' Movie Website Using Enter Pro — aftahi_ai · 2026-08-13