Affordance Segmentation on EPIC-Diff (test)
20.6S01 ScoreDIV-FF
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| DIV-FFModel components=full model, Inference modality=video inference2025.03 | 20.6 | 19.9 | 14.4 | 22.4 | 30.1 | 22.3 | 20.1 | 16.8 | 17.1 | 23.1 | 20.7 | |
| LERF2025.03 | 18.2 | 17.4 | 6.8 | 11.5 | 11.9 | 18.4 | 11.7 | 7.5 | 15.2 | 4.2 | 12.2 | |
| DIV-FFModel components=full model, Inference modality=image inference2025.03 | 17.3 | 13.7 | 6.2 | 13.7 | 19.1 | 8.1 | 18.5 | 7.1 | 11.1 | 3.6 | 11.8 | |
| DIV-FFFeature granularity=CLIP in patches2025.03 | 17.1 | 15.6 | 7.1 | 9.4 | 12.9 | 19.7 | 12.4 | 11.3 | 15.3 | 12.6 | 13.3 | |
| OWL-VIT2025.03 | 4.8 | 4.2 | 1.4 | 5.6 | 4.8 | 2.3 | 13.2 | 4 | 5.4 | 4.4 | 5 | |
| OWL-VIT + SAMModel components=OWL-VIT with Segment Anything Model (SAM)2025.03 | 4.6 | 5.4 | 1.9 | 4.6 | 5.8 | 1.1 | 8.6 | 4.5 | 7.7 | 8.3 | 5.3 |