A newer version of the Gradio SDK is available: 6.15.2
metadata
title: Egg Damage Top 3 Classifier
sdk: gradio
app_file: app.py
pinned: false
license: mit
Egg Damage Top 3 Classifier
Binary egg image classifier for:
- Damaged
- Not Damaged
This Space deploys the top three trained models from the local evaluation run:
| Rank | Model | Test F1 | ROC-AUC | Accuracy |
|---|---|---|---|---|
| 1 | MobileNetV3 | 0.9674 | 0.9832 | 0.9500 |
| 2 | LBP + SVM | 0.9290 | 0.9394 | 0.8917 |
| 3 | Xception | 0.9231 | 0.9617 | 0.8833 |
Upload a clear egg image, choose a model, and run classification. CNN models can also show a Grad-CAM-style explanation image.