Semantic Segmentation on South Asian Megacities unseen cities 20% labeled source data
0.452mIoUSLUM-i
Evaluation results
| Method | Links | |
|---|---|---|
| SLUM-iBackbone=DINOv22026.02 | 0.452 | |
| UniMatch-v2Backbone=DINOv22026.02 | 0.443 | |
| FixMatchBackbone=ResNet-1012026.02 | 0.44 | |
| UniMatchBackbone=ResNet-1012026.02 | 0.435 | |
| SLUM-iBackbone=ResNet-1012026.02 | 0.431 | |
| SupervisedBackbone=DINOv22026.02 | 0.387 | |
| SupervisedBackbone=ResNet-1012026.02 | 0.326 |