PlantCare AI - efficientnet_b4
Fine-tuned efficientnet_b4.ra2_in1k on the PlantCity field dataset (12 crops, 52 classes).
| metric | value |
|---|---|
| test accuracy | 0.9969 |
| test macro-F1 | 0.9960 |
| ECE (calibrated) | 0.0018 |
| input size | 320x320 |
Split: group-aware stratified 72/14/14, near-duplicates removed, original photos only, on-the-fly augmentation.
Divide logits by temperature (see plantcare_meta.json) before softmax for calibrated confidences.
Limitations: trained on leaves from two districts of Khyber Pakhtunkhwa (Charsadda, Chitral); expect a drop on other regions, cameras and crops. Use as decision support, not a diagnosis.
Dataset: Khan et al. (2025) PlantCity: A Comprehensive Image Based on Multi Crop Leaves disease in Pakistan, Mendeley Data (check its licence before releasing weights).
- Downloads last month
- 3