README.md

metadata

title: Real-time Smart Waste Classification
emoji: ♻️
colorFrom: green
colorTo: blue
sdk: gradio
sdk_version: 5.45.0
app_file: app.py
pinned: false
license: mit
short_description: This model can detect different types of wastes in Real-Time

♻️ Smart Waste Classification

An AI-powered web application that classifies different types of waste materials to help with proper recycling and disposal.

Features

• Image Classification: Upload images of waste items for automatic classification
• 10 Categories: Supports classification of biological waste, glass (brown/green/white), cardboard, metal, paper, plastic, shoes, and general trash
• Recycling Guidance: Provides clear guidance on whether items are recyclable or not
• Statistics Dashboard: Track classification statistics with visual charts
• Export Reports: Download classification data as CSV or PDF reports

Supported Waste Categories

Recyclable ♻️

• Brown Glass
• Green Glass
• White Glass
• Metal
• Plastic
• Paper
• Cardboard

Non-Recyclable 🗑️

• Biological Waste
• Shoes
• General Trash

How to Use

1. Upload Image: Click "Choose File" and select an image of waste material
2. Classify: Click "Classify Waste" to get AI prediction
3. View Result: See the predicted category, confidence score, and recycling guidance
4. Check Stats: Visit the statistics page to see classification trends
5. Export Data: Download your classification history as CSV or PDF

Technology Stack

• Backend: Flask (Python web framework)
• AI Model: EfficientNet-B0 trained on waste classification dataset
• Frontend: Bootstrap 5 for responsive UI
• Data Visualization: Matplotlib for statistics charts
• Export: CSV and PDF report generation

Model Information

The application uses a fine-tuned EfficientNet-B0 model trained specifically for waste classification. The model achieves high accuracy across all supported categories and provides confidence scores for each prediction.

Limitations

• Camera functionality is not available in Hugging Face Spaces
• Maximum file size limits may apply
• Supported image formats: PNG, JPG, JPEG, GIF, BMP

License

This project is licensed under the MIT License.

Acknowledgments

• Built with TensorFlow/Keras
• UI components from Bootstrap
• Background images from Unsplash