Action-based video object segmentation on ActiSeg-NL Name-60, mask clean
60.2p-mIoUCo-teaching
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Co-teachingNoise Type=Name, Noise Level=60%, Mask Condition=clean2025.09 | 60.2 | 13.8 | 68.8 | 23 | 71.2 | 80.9 | |
| Co-teachingVisual Backbone=ResNet-101, Text Encoder=RoBERTa, Noise condition=Name-60, Mask condition=Clean2025.09 | 60.2 | 13.8 | 68.8 | 23 | 71.2 | 80.9 | |
| SCENoise Type=Name, Noise Level=60%, Mask Condition=clean2025.09 | 58.3 | 12.8 | 67.1 | 18.7 | 70.7 | 80.1 | |
| SCEVisual Backbone=ResNet-101, Text Encoder=RoBERTa, Noise condition=Name-60, Mask condition=Clean2025.09 | 58.3 | 12.8 | 67.1 | 18.7 | 70.7 | 80.1 | |
| APLNoise Type=Name, Noise Level=60%, Mask Condition=clean2025.09 | 57.8 | 11.6 | 66.3 | 15.8 | 70.9 | 80.3 | |
| APLVisual Backbone=ResNet-101, Text Encoder=RoBERTa, Noise condition=Name-60, Mask condition=Clean2025.09 | 57.8 | 11.6 | 66.3 | 15.8 | 70.9 | 80.3 | |
| GCENoise Type=Name, Noise Level=60%, Mask Condition=clean2025.09 | 56.7 | 11.5 | 65.3 | 15 | 70 | 81.1 | |
| GCEVisual Backbone=ResNet-101, Text Encoder=RoBERTa, Noise condition=Name-60, Mask condition=Clean2025.09 | 56.7 | 11.5 | 65.3 | 15 | 70 | 81.1 | |
| ELRNoise Type=Name, Noise Level=60%, Mask Condition=clean2025.09 | 48.5 | 9.8 | 51.8 | 11.8 | 66.7 | 72.2 | |
| ELRVisual Backbone=ResNet-101, Text Encoder=RoBERTa, Noise condition=Name-60, Mask condition=Clean2025.09 | 48.5 | 9.8 | 51.8 | 11.8 | 66.7 | 72.2 | |
| NPNNoise Type=Name, Noise Level=60%, Mask Condition=clean2025.09 | 47.6 | 9.2 | 51.6 | 7.9 | 66.3 | 71.8 | |
| NPNVisual Backbone=ResNet-101, Text Encoder=RoBERTa, Noise condition=Name-60, Mask condition=Clean2025.09 | 47.6 | 9.2 | 51.6 | 7.9 | 66.3 | 71.8 |