Google has introduced a new AI model named SignGemma, designed to convert sign language into text in real-time. This innovation was announced at the Google I/O 2025 event and is currently in the testing phase, with public release expected by the end of the year.
SignGemma: A Leap in Sign Language Translation
SignGemma is the most advanced model by Google that understands sign language gestures, facial expressions, and body movements. It can accurately translate American Sign Language (ASL) into English text and will expand to other languages in the future.
According to Google, this AI model processes visual cues from the signer and turns them into readable text, helping bridge the communication gap for the deaf and hard-of-hearing community.
In a move toward open innovation, Google plans to release SignGemma as open-source, allowing developers and researchers worldwide to build accessible tools and applications for sign language communication.
Google’s Past Efforts: SignTown and Project Shuwa
Before SignGemma, Google launched an educational tool called SignTown, an interactive web app that helps users learn sign language by recognizing hand gestures and expressions through a webcam.
Additionally, Google’s Project Shuwa (named after the Japanese word for “sign language”) aims to improve search results and build sign language dictionaries to enhance accessibility across the web.
A Vision for Inclusive Technology
SignGemma represents a significant step toward making technology more inclusive. With tools like these, individuals with hearing disabilities can participate more fully in education, workplaces, and social interactions.
By making this technology open and widely accessible, Google sets a new standard for tech companies to prioritize accessibility and social impact.
