Inspiration

What it does

How we built it

SignBridge 🤟

Real-time sign language translation for video calls. Communication without barriers.

SignBridge is an AI-powered video calling platform that detects hand gestures and converts them into live text captions. Built to make video communication accessible for the 466 million deaf and hard-of-hearing people worldwide.

Features

• 🎯 Real-time gesture detection (10 FPS, <100ms latency)
• 💬 Live caption overlay on video feeds
• 🔒 Privacy-first: All processing happens in your browser
• 🚀 Zero setup: No accounts, no downloads
• 🌐 WebRTC peer-to-peer video streaming
• ⚡ Built with React, TensorFlow.js, and HandPose AI

Quick Start

1. Start the server: cd server && node server.js
2. Start the client: npm run dev
3. Share room code with a friend
4. Start signing!

Impact

Making healthcare, education, workplaces, and social connections accessible for everyone.

Challenges we ran into

Accomplishments that we're proud of

We made and end to end project that can actually track real time hand movements to give instant replies.

What we learned

we learned working with machine learning models and working in AI/ML while also making a full stack app

What's next for SignBridge

next step would be making a full block custom machine learning model which can be trained tknow everything about hand sign language .

Built With

• git
• javascript
• node.js
• numpy
• python
• react
• tensorflow