DeepMind has developed a new model called sign-language-to-text (SL2T). This model aims to enable sign language recognition and transcription.
SL2T is intended to power features that assist Deaf and hard of hearing users by converting sign language into readable text.
The announcement highlights the potential for sign language AI to improve accessibility and communication. Details on model size, architecture, or performance metrics are not specified.
Source: https://deepmind.google/blog/putting-sign-language-ai-into-users-hands/