Inspiration
Helping People to communicate with those that use sign language as a bridge for communication.
What it does
Introducing SignBridge, an AI-powered edge-based sign language interpreter designed to break down communication barriers for the deaf and hard-of-hearing community. Using a mobile device or IoT camera, SignBridge streams live video of sign language gestures, processes it locally using AWS IoT Greengrass for low-latency inference, and sends frame-level data to Amazon Bedrock for accurate real-time interpretation. The interpreted content is instantly displayed as on-screen text or converted to spoken audio, enabling seamless communication in classrooms, hospitals, workplaces, and public spaces—even in environments with limited connectivity.
SignBridge leverages the power of AWS’s generative AI and edge services to create an inclusive and responsive communication assistant. With support for 5G and multilingual translation powered by AWS Q, the system ensures global accessibility while respecting privacy by keeping data at the edge whenever possible. Whether helping a student participate in real-time discussions or enabling a traveler to ask for help in an unfamiliar language, SignBridge is a transformative tool for accessibility and empowerment.## How we built it
Challenges we ran into
Issues with camera access took time to resolve.
Accomplishments that we're proud of
Meeting the tight project deadline.
What we learned
How to deploy this new stack in AWS.
What's next for Sign Bridge - Personal Sign Language Translation
Perhaps a mobile app...
Built with
Amazon Bedrock (Claude) – Real-time sign interpretation via multimodal LLM AWS IoT + Greengrass – Local, edge-optimized inference for low latency Amazon Q – Optional multilingual and contextual translation Amazon S3 + API Gateway + Lambda – Backend support for session data and integrations
Built With
- amazon-web-services
- bedrock
- claude
- javascript
Log in or sign up for Devpost to join the conversation.