Traversability Estimation on ACDC
74.7IoUViTA
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| ViTAType=VFM (vision foundation model-based)2026.05 | 74.7 | 78.7 | 91.1 | |
| SAM2 FTType=VFM (vision foundation model-based), Evaluation Protocol=FT (full fine-tuning)2026.05 | 70.9 | 71.3 | 98.9 | |
| GeNIEType=VFM (vision foundation model-based)2026.05 | 70.8 | 71.1 | 97.3 | |
| ST-SegType=SS (semantic segmentation)2026.05 | 69.3 | 69.7 | 99.2 | |
| Mask2FormerType=SS (semantic segmentation)2026.05 | 68.8 | 69.7 | 98.1 | |
| SAM2 FZType=VFM (vision foundation model-based), Evaluation Protocol=FZ (encoder frozen)2026.05 | 68.7 | 69.7 | 97.9 | |
| SegFormerType=SS (semantic segmentation)2026.05 | 64.7 | 65.9 | 97.2 | |
| DINOv3 FTType=VFM (vision foundation model-based), Evaluation Protocol=FT (full fine-tuning)2026.05 | 60.3 | 63.6 | 93.8 | |
| GANavType=SS (semantic segmentation)2026.05 | 53.7 | 62.8 | 78.8 | |
| DINOv3 FZType=VFM (vision foundation model-based), Evaluation Protocol=FZ (encoder frozen)2026.05 | 48.7 | 50.8 | 92.3 | |
| STEPPType=SELF (self-supervised traversability)2026.05 | 36.4 | 66.1 | 44.7 |