Indian Sign Language Recognition(ISLAR)
summary
Claims an AI system that converts regional ISL into text and speech in real time, developed with the Centre for Research and Development of Deaf & Mute and Google and tested across Mumbai/Pune variants. Uses TensorFlow, Vision API, MediaPipe and Google text‑to‑speech; starts with color‑glove color segmentation then moves to skeletal and facial keypoints, aggregates 10‑second frame windows, trains on thousands of photos, covers ~20–25 phrases, targets CPU‑only mobile deployment after 18–24 months of work, and reports promising but imperfect accuracy requiring sequential models for full sentences.
AI-written summary · about
chapters (9)
- 0:00Introduction to Indian Sign Language Recognition
- 6:11Motivation for ISLAR
- 8:18Challenges of Indian Sign Language
- 10:48Sign language recognition system
- 13:10Facial expressions and gestures
- 16:44Google collaboration & future plans
- 20:46Contribute to ISLAR awareness
- 22:37Personal projects & passions
- 23:42Q&A
related talks
08mIndian Sign Language Recognition (ISLAR)
24mSIGN TO TEXT Unleashing the Power of Google AI for real time ASL transcription
14mAI for a Billion Voices: Taking AI where there is Maximum Impact
46mArtificial intelligence to the rescue of accessibility
27mSign Language Recognition Improvements Since the London Keynote
31mDigital Accessibility: The First Artificial Human That Knows Italian Sign Language
09mBeyond Captions A Practical Pipeline for Live AI-Generated Sign Language Avatars
36mSession on 'Learning to learn'