A smart waste-sorting assistant that uses computer vision to identify and classify food tray contents into trash, recycle, compost, or dish return — making disposal effortless and eco-friendly.

• Real-time object detection using YOLOv11, the latest version of YOLO
• Webcam integration for live detection using Gemini 2.0
• Image upload functionality for analyzing static images
• Focused detection of food and dining-related items
• Configurable confidence threshold for detections

• React: A JavaScript library for building user interfaces.
• TypeScript: A strongly typed superset of JavaScript.
• React Router: For navigation and routing between pages.
• Tailwind CSS: For styling components with utility-first CSS classes.
• Axios: For making HTTP requests to the backend.

• Flask: A Python web framework for building the backend API.
• YOLO (Ultralytics): For object detection.
• Google Gemini: For AI object identification and classification.
• OpenCV for image processing and webcam access

• Auth0

• Node.js 16.0+: Required for running the frontend.
• Python 3.8+: Required for running the backend.
• Package Managers:
  • npm or yarn for the frontend.
  • pip for the backend.

1. Clone the repository and open it locally:

   git clone https://github.com/F4llow/trayce.git
   cd trayce
   

2. Open the backend folder and install the required packages inside of a virtual environment:

   cd backend
   python3 -m venv venv
   source venv/bin/activate
   pip install -r requirements.txt
   

3. Add a environment file named .env and assign your Gemini API key to GEMINI_API_KEY.

4. Run the application's backend:

   python3 app.py
   

• Click Login and sign in with Gmail

• Click on the "Start Scanning" button to start the webcam.

• Make sure that the webcam is under good lighting and that all items are clearly visible.
• Click on the "Capture Image" button to take a picture of your tray.

• View your tray analysis as an image labeled by Gemini AI object identification.
• Click the "View All Items" button to view the full list of identified objects, sorted by category.
• Click the "Scan Again" button to restart the process.

This project is licensed under the MIT License - see the LICENSE file for details.

• Google Gemini's API for computer vision object classification.