Classification on Office-Home
84Average AccuracyOURS
Evaluation results
| Method | Links | ||
|---|---|---|---|
| OURSBackbone=ConvNeXt-B, Learning rate=1×10-4, Batch size=4, Weight decay=1×10-4, Consistency loss weight (λ)=0.52026.01 | 84 | 8 | |
| DIFO-C-B322026.01 | 83.1 | — | |
| GMM-DaDiLBackbone=ResNet-101, Access to source labels (X x Y)=None2025.10 | 78.81 | — | |
| WGFBackbone=ResNet-101, Access to source labels (X x Y)=Yes2025.10 | 78.42 | — | |
| DaDiL-EBackbone=ResNet-101, Access to source labels (X x Y)=None2025.10 | 78.14 | — | |
| GMMBackbone=ResNet-101, Access to source labels (X x Y)=Yes2025.10 | 77.87 | — | |
| DaDiL-RBackbone=ResNet-101, Access to source labels (X x Y)=None2025.10 | 77.86 | — | |
| WGFBackbone=ResNet-101, Access to source labels (X x Y)=No2025.10 | 77.56 | — | |
| DiscreteBackbone=ResNet-101, Access to source labels (X x Y)=No2025.10 | 77.17 | — | |
| DiscreteBackbone=ResNet-101, Access to source labels (X x Y)=Yes2025.10 | 77.09 | — | |
| U-NOTBackbone=ResNet-101, Access to source labels (X x Y)=No2025.10 | 76.85 | — | |
| NormFlowBackbone=ResNet-101, Access to source labels (X x Y)=No2025.10 | 76.73 | — | |
| WJDOTBackbone=ResNet-101, Access to source labels (X x Y)=None2025.10 | 76.59 | — | |
| CW2BBackbone=ResNet-101, Access to source labels (X x Y)=No2025.10 | 76.44 | — | |
| Source-OnlyBackbone=ResNet-101, Access to source labels (X x Y)=None2025.10 | 75.95 | — | |
| NOTBackbone=ResNet-101, Access to source labels (X x Y)=No2025.10 | 75.36 | — | |
| DualTTA2026.04 | 61.51 | — | |
| Tent2026.04 | 61.08 | — | |
| SAR2026.04 | 59.82 | — | |
| No adapt2026.04 | 59.14 | — | |
| DeYO2026.04 | 59.08 | — | |
| EATA2026.04 | 52.34 | — |