Google Research has developed StreetReaderAI, a prototype system designed to make Google Street View accessible to individuals who are blind or have low vision. The system utilizes multimodal AI, specifically Google's Gemini models, to provide real-time audio descriptions of street scenes and allow for conversational interaction about the environment. Users can navigate through virtual streetscapes using keyboard commands or voice input, receiving dynamic audio feedback about their surroundings and potential points of interest. AI
Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →
RANK_REASON The item describes a prototype system developed by Google Research, which is a significant step in accessibility but not a widely released product or a frontier model release.
