Semantic Segmentation on MVSeg Sequential (INO, KAIST, RGBT234) 15
32.6mIoU (INO)DIGA
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| DIGABackbone=ResNet-101, Model=DeepLabV3, Backpropagation=false2026.07 | 32.6 | 33.6 | 31.6 | 32.6 | |
| TestMate + DIGABackbone=ResNet-101, Model=DeepLabV3, Backpropagation=false2026.07 | 32 | 36.8 | 32.3 | 33.7 | |
| TENTBackbone=ResNet-101, Model=DeepLabV3, Backpropagation=true2026.07 | 29.5 | 35.6 | 31.9 | 32.3 | |
| TestMateBackbone=ResNet-101, Model=DeepLabV3, Backpropagation=false2026.07 | 29 | 35 | 33.8 | 32.6 | |
| TestMate + CoTTABackbone=ResNet-101, Model=DeepLabV3, Backpropagation=true2026.07 | 28.4 | 34.8 | 33.6 | 32.3 | |
| SourceBackbone=ResNet-101, Model=DeepLabV3, Backpropagation=false2026.07 | 27.9 | 34.1 | 32.3 | 31.4 | |
| EATABackbone=ResNet-101, Model=DeepLabV3, Backpropagation=true2026.07 | 27.9 | 36.2 | 32.2 | 32.1 | |
| CoTTaBackbone=ResNet-101, Model=DeepLabV3, Backpropagation=true2026.07 | 27.5 | 35.9 | 32.1 | 31.8 |