| language: en | |
| license: apache-2.0 | |
| library_name: tensorflow | |
| pipeline_tag: image-classification | |
| tags: | |
| - waste-classification | |
| - efficientnet | |
| - recycling | |
| - sustainability | |
| # Waste Classifier (EfficientNetB0) | |
| Image classification model for waste segregation into three classes: | |
| - dry_waste | |
| - wet_waste | |
| - other_waste | |
| ## Model Files | |
| - waste_classifier.keras: TensorFlow Keras model artifact | |
| - inference.py: Local inference utility | |
| - requirements.txt: Python dependencies for local inference | |
| ## Quick Start | |
| ```python | |
| from inference import predict | |
| result = predict("sample.jpg") | |
| print(result) | |
| ``` | |
| ## Output Format | |
| ```json | |
| { | |
| "class": "dry_waste", | |
| "display_name": "♻️ Dry Waste", | |
| "confidence": 92.5, | |
| "action": "Can be recycled → Paper, plastic, metal recovery", | |
| "all_predictions": { | |
| "dry_waste": 92.5, | |
| "other_waste": 4.1, | |
| "wet_waste": 3.4 | |
| } | |
| } | |
| ``` | |