Iris — AI Environment Detector

Visit Website

Overview

Iris is an AI-powered environment detector that uses computer vision and audio processing to detect objects and colors in real-time. It leverages various AI models to provide accurate and efficient detection capabilities.

Features

• Real-time Object Detection: Detects objects using pre-trained models.
• Color Detection: Identifies colors in the video feed.
• Text-to-Speech: Converts detected objects and colors into speech.
• Camera Switching: Switch between front and rear cameras.
• Localization: Supports multiple languages for UI and output.

Usage

1. optional: serve /web from ordinary http server
2. Open index.html in your browser.
3. Allow camera access when prompted.
4. Use the buttons to toggle object detection, color detection, and audio output.

File Structure

• web/index.html: Main HTML file.
• web/index.js: Main JavaScript file for handling logic.
• web/localisation.js: Handles localization.
• web/manifest.json: PWA manifest file.
• web/index.css: Main CSS file.

Acknowledgements

• yolov10n is used for object detection.
• Xenova/speecht5_tts for the text-to-speech pipeline.
• Tracking.js for color detection.