EyeHear

What it does

EyeHear uses the Gemini API to make doorbell cameras accessible to sight-impaired individuals. Gemini generates detailed text descriptions of doorbell videos using multimodal prompting. These text descriptions are then converted to audio which is auto-played in a browser. This allows users to "hear" what's at the door, including identifying people, packages, or other events. Video description metadata is stored in Firebase for history and possible analytics purposes.