Eyedentify is a wearable face-recognition-enabled cap for the blind, developed to offer real-time audio feedback on nearby people’s identities using AI-based image processing. A camera fitted on a cap captures video of nearby faces. A Raspberry Pi-based system compares the detected face to stored images. If a match is found, an audio cue with the person’s name is played through a speaker. If no match is found, a default sound is played to indicate "unknown person."  

Origin of Idea Inspired by Mr. Didwania, a blind individual who relied on his wife to whisper names of people during walks. The device was conceived to help him independently recognize people. Constraints Must work across multiple languages (English, Hindi, Gujarati) Should be non-intrusive in design Needs to be portable and lightweight Must be usable without external assistance Brainstormed Concepts Cap-Based Device Uses camera + embedded ML for real-time facial recognition Provides audio descriptions via speaker Tactile Remote-Like Device Features buttons with different textures Functions include distance sensing, face recognition, and SOS alert Final Prototype Camera sends real-time feed to Raspberry Pi Facial features are compared with stored images Based on match, system plays corresponding audio file via HXJ8002 amplifier & speaker Enclosed in a wearable cap with minimal external wiring Images: Skills Applied CAD | Electronics | Design Thinking | Programming | Soldering | Prototyping