Zero-Shot Video Object Segmentation on 17 video datasets
79.3Zero-shot J&F AccuracySAM 2.1
Evaluation results
| Method | Links | |
|---|---|---|
| SAM 2.1Prompt=ground-truth mask2025.01 | 79.3 | |
| SAM 2Prompt=ground-truth mask2025.01 | 77.6 | |
| SAM 2.1Prompt=5-click2025.01 | 77.6 | |
| EdgeTAMPrompt=ground-truth mask2025.01 | 77 | |
| EdgeTAMPrompt=5-click2025.01 | 75.5 | |
| SAM 2Prompt=5-click2025.01 | 75.4 | |
| SAM 2.1Prompt=3-click2025.01 | 75.3 | |
| SAM 2.1Prompt=bounding box2025.01 | 74.4 | |
| SAM + CutiePrompt=ground-truth mask2025.01 | 74.1 | |
| SAM 2Prompt=3-click2025.01 | 73.2 | |
| SAM 2Prompt=bounding box2025.01 | 72.9 | |
| SAM + XMem++Prompt=ground-truth mask2025.01 | 72.7 | |
| EdgeTAMPrompt=3-click2025.01 | 72.7 | |
| SAM + CutiePrompt=5-click2025.01 | 72.2 | |
| EdgeTAMPrompt=bounding box2025.01 | 71.3 | |
| SAM + XMem++Prompt=5-click2025.01 | 70.6 | |
| SAM + CutiePrompt=3-click2025.01 | 70.1 | |
| SAM + CutiePrompt=bounding box2025.01 | 69.4 | |
| SAM + XMem++Prompt=3-click2025.01 | 68.4 | |
| SAM + XMem++Prompt=bounding box2025.01 | 67.6 | |
| SAM 2.1Prompt=1-click2025.01 | 64.7 | |
| SAM 2Prompt=1-click2025.01 | 64.3 | |
| SAM + XMem++Prompt=1-click2025.01 | 56.9 | |
| SAM + CutiePrompt=1-click2025.01 | 56.7 | |
| EdgeTAMPrompt=1-click2025.01 | 54.4 |