Semantic Segmentation on Cityscapes Adverse (test)
71.8mIoU (Rainy)SETR
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| SETR2026.02 | 71.8 | 70.6 | 73.2 | 71.3 | 61.4 | 61.9 | 70.3 | 73 | 68 | |
| ICNet2026.02 | 68.8 | 69.3 | 70.6 | 66.7 | 27 | 54.7 | 61.2 | 67.5 | 61 | |
| Mask2Former2026.02 | 67.7 | 72.9 | 74.9 | 73.7 | 62.8 | 63 | 69 | 75 | 68.2 | |
| SegFormerTraining Modality=RGB, Test Modality=RGB2026.02 | 65.9 | 65.7 | 69 | 65.4 | 46.9 | 57.9 | 63.4 | 66.1 | 61.2 | |
| DDRNet2026.02 | 65.6 | 69.3 | 71.9 | 55.1 | 30.2 | 53.5 | 58.9 | 64.8 | 59.3 | |
| SegFormer_L2Variant=L2 alignment baseline, Training Modality=RGB+E, Test Modality=RGB2026.02 | 63.5 | 66.9 | 69.1 | 66.2 | 46.3 | 57.8 | 63.2 | 67.5 | 62.5 | |
| SegFormer_PEPRVariant=PEPR (Privileged Event-based Predictive Regularization), Training Modality=RGB+E, Test Modality=RGB2026.02 | 62.8 | 67.8 | 70.2 | 66.6 | 47.1 | 59 | 63.9 | 67.7 | 63.1 | |
| DeepLabV3+2026.02 | 58.2 | 66.1 | 69.5 | 46.8 | 22.3 | 42.6 | 47.4 | 57.7 | 51.2 |