Google has unveiled SignGemma, a novel artificial intelligence (AI) model designed to convert sign language into spoken text. This model, which will be included in the Gemma series, is currently undergoing testing by the technology giant based in Mountain View and is anticipated to be launched later this year. This blog discusses about Google Unveils SignGemma.

Like the other models in the Gemma series, SignGemma will also be an open-source AI model, accessible to both individuals and businesses. It was initially presented during the Google I/O 2025 keynote and aims to assist individuals with speech and hearing disabilities in effectively communicating with those who may not understand sign language.

SignGemma Can Track Hand Movements and Facial Expressions

In a post on X, the official account of Google DeepMind shared a demonstration of the AI model along with some information regarding its release date. However, this is not the first instance of SignGemma being showcased. It was also briefly presented at the Google I/O event by Gus Martin, the Gemma Product Manager at DeepMind. This article enumerates about Google Unveils SignGemma.

During the presentation, Martin emphasized that the AI model is capable of delivering real-time text translation from sign language, facilitating seamless face-to-face communication. The model has been trained on datasets encompassing various styles of sign languages, although it performs optimally with American Sign Language (ASL) when translating into English.

Summary

As reported by MultiLingual, due to its open-source nature, SignGemma can operate without requiring an Internet connection. This feature makes it particularly useful in regions with limited connectivity. It is said to be constructed on the Gemini Nano framework and employs a vision transformer to monitor and analyze hand movements, shapes, and facial expressions. In addition to making it available to developers, Google may incorporate the model into its existing AI tools, such as Gemini Live. Describing it as “our most capable model for translating sign language into spoken text,” DeepMind has indicated that it will be released later this year.

LEAD GENERATION SERVICES APPSREAD