Table 1 Comparison of five model accuracy metrics.
From: Landslide detection using multimodal data fusion and an improved Deeplabv3+ model
Model | Precision | Accuracy | Recall | F1-Score | IoU (%) |
|---|---|---|---|---|---|
UNet | 0.770 | 0.845 | 0.775 | 0.773 | 72.9 |
DeepLabv3+ | 0.810 | 0.855 | 0.815 | 0.813 | 76.5 |
Swin Transformer | 0.831 | 0.872 | 0.831 | 0.833 | 78.6 |
SegFormer | 0.830 | 0.875 | 0.835 | 0.832 | 78.9 |
HRNet | 0.815 | 0.875 | 0.845 | 0.830 | 77.6 |
Fast-SCNN | 0.748 | 0.843 | 0.782 | 0.764 | 69.2 |
FCA-DeepLab | 0.865 | 0.912 | 0.870 | 0.867 | 81.8 |
