SignBridge AI

SignBridge AI — Every Gesture Deserves a Voice SignBridge AI is an AI-powered accessibility application designed to break communication barriers between Deaf people who use sign language and those who do not understand it. Our project focuses on Dominican Sign Language (LSRD), an underrepresented language in artificial intelligence. Using computer vision and machine learning, SignBridge AI recognizes a limited set of everyday sign expressions and transforms them into spoken communication in Spanish or English. The application combines MediaPipe for real-time hand tracking, a custom-trained temporal recognition model for identifying sign sequences, and a Go backend integrated with Meta's Muse Spark through the Meta Model API. Muse Spark provides contextual language interpretation, while browser-based speech synthesis converts the resulting messages into audible speech. The current prototype focuses on practical communication scenarios, including greetings, requesting assistance, asking for a restroom, expressing gratitude, and requesting medical help. Recognition is experimental and limited to trained expressions rather than unrestricted sign language translation. SignBridge AI demonstrates how computer vision and generative AI can work together to make everyday communication more accessible, particularly for underserved linguistic communities. Our vision is simple: a world where communication barriers become smaller, one gesture at a time.