LAI Estimation on Winter wheat (test)
0.8123R2MCVI-SANet (Finetuned)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MCVI-SANet (Finetuned)Training strategy=Finetuned2025.12 | 0.8123 | 0.4796 | |
| MCVI-SANet (Unfrozen)Training strategy=Unfrozen2025.12 | 0.807 | 0.4861 | |
| ResNet18Model category=Baseline deep learning2025.12 | 0.7456 | 0.5585 | |
| MCVI-SANet (VI-SABlock→CBAM)Attention mechanism=CBAM, Model Category=Our models2025.12 | 0.7429 | 0.5568 | |
| MCVI-SANet (VI-SABlock→ECA)Attention mechanism=ECA, Model Category=Our models2025.12 | 0.7324 | 0.5698 | |
| MCVI-SANet (w/o VI-SABlock)Configuration=Ablation without VI-SABlock2025.12 | 0.7316 | 0.5723 | |
| SVRInput features=VIs+TF, Model category=Traditional machine learning2025.12 | 0.7278 | 0.5782 | |
| MCVI-SANet (VI-SABlock→SE)Attention mechanism=SE, Model Category=Our models2025.12 | 0.7198 | 0.5825 | |
| XGBoostInput features=VIs+TFs, Model category=Traditional machine learning2025.12 | 0.692 | 0.6149 | |
| SVRInput features=VIs, Model category=Traditional machine learning2025.12 | 0.6855 | 0.6216 | |
| RFRInput features=VIs+TFs, Model category=Traditional machine learning2025.12 | 0.6835 | 0.623 | |
| MobileNet V3Model category=Baseline deep learning2025.12 | 0.677 | 0.6278 | |
| ResNet50Model category=Baseline deep learning2025.12 | 0.6703 | 0.6335 | |
| RFRInput features=VIs, Model category=Traditional machine learning2025.12 | 0.6677 | 0.6388 | |
| EfficientNet B0Model category=Baseline deep learning2025.12 | 0.6603 | 0.6445 | |
| MobileNet V2Model category=Baseline deep learning2025.12 | 0.634 | 0.6681 | |
| XGBoostInput features=VIs, Model category=Traditional machine learning2025.12 | 0.6265 | 0.677 | |
| ShuffleNet V2Model category=Baseline deep learning2025.12 | 0.5684 | 0.7253 | |
| EfficientNet V2-SModel category=Baseline deep learning2025.12 | 0.5553 | 0.737 | |
| PLSRInput features=VIs+TFs, Model category=Traditional machine learning2025.12 | 0.5206 | 0.7673 | |
| PLSRInput features=VIs, Model category=Traditional machine learning2025.12 | 0.5103 | 0.7756 |