Table 1 Results of ablation experiments.
From: GLE-net: global-local information enhancement for semantic segmentation of remote sensing images
Method | Model | IoU\(\%\) | Evaluation Index\(\%\) | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
MFM | SIEM | FCM | Imp.surf. | Buliding | Lowveg. | Tree | Car | MIoU | Mean.F1 | OA | |
ST | 78.54 | 82.20 | 75.66 | 79.02 | 77.15 | 81.35 | 75.88 | 79.11 | |||
ST+MFM | \(\checkmark\) | 79.32 | 83.66 | 76.98 | 82.06 | 78.855 | 83.85 | 77.99 | 80.16 | ||
ST+SIEM | \(\checkmark\) | 78.33 | 83.99 | 76.35 | 81.48 | 78.39 | 83.44 | 78.62 | 81.15 | ||
ST+FCM | \(\checkmark\) | 78.62 | 83.08 | 76.28 | 82.28 | 78.73 | 83.69 | 78.97 | 81.09 | ||
ST+MFM+SIEM | \(\checkmark\) | \(\checkmark\) | 79.34 | 84.51 | 76.22 | 82.39 | 79.64 | 84.11 | 77.81 | 82.28 | |
ST+MFM+FCM | \(\checkmark\) | \(\checkmark\) | 79.97 | 84.91 | 77.11 | 82.52 | 78.96 | 83.74 | 77.15 | 82.35 | |
ST+SIEM+FCM | \(\checkmark\) | \(\checkmark\) | 79.08 | 84.86 | 76.97 | 82.56 | 79.67 | 84.89 | 77.89 | 81.78 | |
ST+MFM+SIEM+FCM | \(\checkmark\) | \(\checkmark\) | \(\checkmark\) | 80.64 | 85.98 | 77.33 | 82.55 | 79.89 | 85.03 | 78.05 | 82.46 |
