Semantic Segmentation on SAR Road Extraction Dataset (test)
77.76PrecisionUPerNet
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| UPerNetBackbone=DI3CL-ResNet1012025.11 | 77.76 | 71.7 | 74.6 | 59.5 | |
| UPerNetBackbone=RSP-ResNet502025.11 | 77.28 | 69.23 | 73.03 | 57.52 | |
| PSPNetBackbone=IMP-ResNet502025.11 | 75.95 | 64.93 | 70.01 | 53.86 | |
| UPerNetBackbone=IMP-ResNet502025.11 | 75.78 | 70.54 | 73.07 | 57.56 | |
| UPerNetBackbone=DI3CL-ResNet502025.11 | 74.83 | 72.9 | 73.85 | 58.54 | |
| UPerNetBackbone=SeCo-ResNet502025.11 | 74.82 | 68.32 | 71.42 | 55.55 | |
| UNet2025.11 | 74.14 | 66.16 | 69.92 | 53.75 | |
| FCNBackbone=IMP-ResNet502025.11 | 73.27 | 70.07 | 71.63 | 55.8 | |
| UPerNetBackbone=SARATR-X+HiViT-B2025.11 | 71.78 | 67.45 | 69.55 | 53.31 | |
| DANetBackbone=IMP-ResNet502025.11 | 70.81 | 72.74 | 71.76 | 55.96 | |
| UPerNetBackbone=SAR-JEPA+ViT-B2025.11 | 70.34 | 60.16 | 64.85 | 48 | |
| FarSegBackbone=IMP-ResNet502025.11 | 70.29 | 69.06 | 69.67 | 53.46 | |
| DeepLabV3+Backbone=IMP-ResNet502025.11 | 70.27 | 72.66 | 71.44 | 55.57 |