This project was developed during the Shell-Edunet Skills4Future Internship (June–July 2025). It aims to classify garbage images into six categories using deep learning and transfer learning, achieving up to 98% accuracy.

Automatically classify garbage into:

• Cardboard
• Glass
• Metal
• Paper
• Plastic
• Trash

• Folder: TrashType_Image_Dataset
• Loaded via ImageDataGenerator with train/validation split
• Includes preprocessing and augmentation for better generalization

• Base Models: EfficientNetV2B2 (primary), MobileNetV2 (for comparison)
• Framework: TensorFlow / Keras
• Accuracy Achieved: ✅ 98%

Garbage_Classification/ ├── Week1/ ├── Week2/ ├── Dataset/ └── README.md

• High-accuracy image classification (98%)
• Transfer learning with EfficientNetV2B2
• Deployed via Gradio / Streamlit for live predictions

🔗 GitHub Link