Edge Detection on NYUD Standard Evaluation - SEval v2 (val)
79.4ODSSAUGE
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| SAUGEInput modality=RGB, Training regime=Zero-shot, Train Dataset=BSDS2026.02 | 79.4 | 80.3 | 81.3 | |
| EdgeSAMInput modality=RGB, Training regime=Supervised2026.02 | 78.3 | 79.7 | 80.5 | |
| EDMBInput modality=RGB, Training regime=Supervised2026.02 | 78.3 | 78.9 | — | |
| RankEDInput modality=RGB, Training regime=Supervised2026.02 | 78 | 79.3 | 82.6 | |
| EDTERInput modality=RGB, Training regime=Supervised2026.02 | 77.4 | 78.9 | 79.7 | |
| RankED + MATCHEDInput modality=RGB, Training regime=Supervised, Post-processing=MATCHED2026.02 | 77.4 | 78.4 | 80.2 | |
| Diff.EdgeInput modality=RGB, Training regime=Supervised2026.02 | 76.1 | 76.6 | — | |
| SAUGE + MATCHEDInput modality=RGB, Training regime=Zero-shot, Train Dataset=BSDS, Post-processing=MATCHED2026.02 | 76 | 78.4 | 81.1 | |
| Diff.Edge + MATCHEDInput modality=RGB, Training regime=Supervised, Post-processing=MATCHED2026.02 | 75.8 | 76.2 | 70.3 | |
| SAUGE†Input modality=RGB, Training regime=Zero-shot, Train Dataset=BSDS, Checkpoint=Official2026.02 | 75.8 | 78.2 | 81 | |
| Diff.Edge*Input modality=RGB, Training regime=Supervised, Training Source=Our training2026.02 | 75.5 | 76.2 | 72.3 | |
| BDCNInput modality=RGB, Training regime=Supervised2026.02 | 74.8 | 76.2 | 77.8 | |
| RCFInput modality=RGB, Training regime=Supervised2026.02 | 74.5 | 75.9 | — | |
| LPCBInput modality=RGB, Training regime=Supervised2026.02 | 73.9 | 75.4 | — | |
| PiDiNet + MATCHEDInput modality=RGB, Training regime=Supervised, Post-processing=MATCHED2026.02 | 73.6 | 74.9 | 75.6 | |
| PiDiNetInput modality=RGB, Training regime=Supervised2026.02 | 73.3 | 74.7 | — | |
| HEDInput modality=RGB, Training regime=Supervised2026.02 | 72.2 | 73.7 | — |