SignLens - AI-Based Sign Language Interpreter

A real-time web application that interprets sign language gestures using computer vision and machine learning (MediaPipe). Designed for accessibility and inclusive education.

🚀 Live Demo

**Click here to view the live project

✨ Features

• Real-Time Hand Tracking: Uses MediaPipe Hands for high-performance hand landmark detection.
• Multi-Gesture Recognition: Supports single-hand and two-hand gestures.
• Responsive Design: Works on desktop and mobile browsers.
• Privacy-Focused: All processing happens locally in the browser; video data is never sent to a server.
• Speech-to-Text: Integrated speech recognition for two-way communication.

👐 Supported Gestures

See the full Gesture Guide for detailed instructions.

Basic Numbers

• 0 - 5 (Single hand)
• 6 - 10 (Two hands)

Common Phrases

• Yes / No
• Thumbs Up / Down
• Okay
• Thank You
• Call Me / I Love You
• Hello / Bye
• Help / Stop
• And more...

🛠️ Tech Stack

• Frontend: HTML5, CSS3, JavaScript (Vanilla)
• AI/ML: MediaPipe Hands (Google), Custom Geometric Classifiers
• Deployment: GitHub Pages

📂 Project Structure

D:\hacknova\SignLens\
├───frontend\           # Main application code
│   ├───index.html      # (Old UI restored in root for live site)
│   ├───style.css       # New styling (optional)
│   ├───old-style.css   # Original styling (active)
│   ├───gesture.md      # Guide to gestures
│   └───src\
│       ├───app.js      # Main logic
│       └───gesture-detection.js # Gesture recognition engine
├───backend\            # (Optional backend components)
└───...

👨‍💻 Development

Setup

1. Clone the repository:

   git clone https://github.com/classifiedstudentkabir/Sign-Language-Interpreter.git

2. Navigate to the project directory:

   cd Sign-Language-Interpreter

3. Open index.html in your browser to test locally.

Contributing

1. Pull latest changes: git pull origin main
2. Make your changes.
3. Stage changes: git add .
4. Commit: git commit -m "Description of changes"
5. Push: git push origin main

---

Part of HackNova Hackathon Project - Team SignLens