Domain Adaptation Regression on dSprites
1Transfer Accuracy (C to N)AFN
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| AFNBackbone=ResNet-182023.03 | 1 | 0.96 | 0.16 | 0.62 | 0.08 | 0.32 | 0.523 | |
| Resnet-18Backbone=ResNet-182023.03 | 0.94 | 0.9 | 0.16 | 0.65 | 0.08 | 0.26 | 0.498 | |
| TCABackbone=ResNet-182023.03 | 0.94 | 0.87 | 0.19 | 0.66 | 0.1 | 0.23 | 0.498 | |
| JDOTBackbone=ResNet-182023.03 | 0.86 | 0.79 | 0.19 | 0.64 | 0.1 | 0.23 | 0.468 | |
| MCDBackbone=ResNet-182023.03 | 0.81 | 0.81 | 0.17 | 0.65 | 0.07 | 0.19 | 0.45 | |
| DANBackbone=ResNet-182023.03 | 0.7 | 0.77 | 0.12 | 0.5 | 0.06 | 0.11 | 0.377 | |
| DANNBackbone=ResNet-182023.03 | 0.47 | 0.46 | 0.16 | 0.65 | 0.05 | 0.1 | 0.315 | |
| RSDBackbone=ResNet-182023.03 | 0.31 | 0.31 | 0.12 | 0.53 | 0.07 | 0.08 | 0.237 | |
| DARE-GRAMBackbone=ResNet-182023.03 | 0.3 | 0.2 | 0.11 | 0.25 | 0.05 | 0.07 | 0.164 |