Cross-Domain Few-Shot Segmentation on Deepglobe
49.71mIoU (1-shot)Ours
Evaluation results
| Method | Links | ||
|---|---|---|---|
| OursBackbone=DINOv3-B, Training Approach=Training-free, Task=CD-FSS, Mark=Ours2026.06 | 49.71 | 58.88 | |
| GF-SAMBackbone=DINOv2-B+SAM-B, Training Approach=Training-free, Task=FSS, Mark=NIPS-242026.06 | 49.05 | 55.13 | |
| GF-SAMBackbone=DINOv3-B+SAM-B, Training Approach=Training-free, Task=FSS, Mark=NIPS-242026.06 | 48.75 | 57.86 | |
| MatcherBackbone=DINOv2-B+SAM-B, Training Approach=Training-free, Task=FSS, Mark=ICLR-242026.06 | 47.01 | 50.98 | |
| MatcherBackbone=DINOv3-B+SAM-B, Training Approach=Training-free, Task=FSS, Mark=ICLR-242026.06 | 46.85 | 53.73 | |
| LoECBackbone=DINOv3-B, Training Approach=Training-based, Task=CD-FSS, Mark=CVPR-252026.06 | 44.34 | 53.25 | |
| ISABackbone=ResNet-50, Training Approach=Training-based, Task=CD-FSS, Mark=ICCV-252026.06 | 44.3 | 52.7 | |
| SDRCBackbone=ViT-B, Training Approach=Training-based, Task=CD-FSS, Mark=ICML-252026.06 | 43.15 | 46.83 | |
| ABCDFSSBackbone=ResNet-50, Training Approach=Training-based, Task=CD-FSS, Mark=CVPR-242026.06 | 42.6 | 49 | |
| LoECBackbone=ViT-B, Training Approach=Training-based, Task=CD-FSS, Mark=CVPR-252026.06 | 42.12 | 51.48 | |
| MAUPBackbone=DINOv2-B+SAM-B, Training Approach=Training-free, Task=FSMIS, Mark=MICCAI-252026.06 | 41.5 | 45.06 | |
| DRABackbone=ResNet-50, Training Approach=Training-based, Task=CD-FSS, Mark=CVPR-242026.06 | 41.29 | 50.12 | |
| APMBackbone=ResNet-50, Training Approach=Training-based, Task=CD-FSS, Mark=NIPS-242026.06 | 40.86 | 44.92 | |
| FSSAMBackbone=DINOv2-B+SAM2, Training Approach=Training-based, Task=FSS, Mark=ICML-252026.06 | 40.25 | 46.58 | |
| SSPBackbone=ResNet-50, Training Approach=Training-based, Task=FSS, Mark=ECCV-222026.06 | 40 | 48.68 | |
| DFNBackbone=ViT-B, Training Approach=Training-based, Task=CD-FSS, Mark=ICML-252026.06 | 39.45 | 47.67 | |
| FPTransBackbone=ViT-B, Training Approach=Training-based, Task=FSS, Mark=NIPS-222026.06 | 38.36 | 49.3 | |
| PATNetBackbone=ResNet-50, Training Approach=Training-based, Task=CD-FSS, Mark=ECCV-222026.06 | 37.89 | 42.97 | |
| PerSAMBackbone=SAM-B, Training Approach=Training-free, Task=FSS, Mark=ICLR-242026.06 | 36.08 | 40.65 | |
| APSegBackbone=SAM-B, Training Approach=Training-based, Task=CD-FSS, Mark=CVPR-242026.06 | 35.94 | 39.98 | |
| PATNetBackbone=DINOv3-B, Training Approach=Training-based, Task=CD-FSS, Mark=ECCV-222026.06 | 33.78 | 40.45 | |
| ABCDFSSBackbone=DINOv3-B, Training Approach=Training-based, Task=CD-FSS, Mark=CVPR-242026.06 | 32.57 | 39.26 | |
| HDMNetBackbone=ResNet-50, Training Approach=Training-based, Task=FSS, Mark=CVPR-232026.06 | 25.4 | 39.1 |