Unsupervised Domain Adaptation on ImageCLEF-DA ImageNet-Pascal-Caltech (test)
77.2Accuracy (I -> P)rRevGrad+CAT
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| rRevGrad+CATBackbone=ResNet-502019.03 | 77.2 | 91 | 95.5 | 91.3 | 75.3 | 93.6 | 87.3 | |
| JANBackbone=ResNet-502019.03 | 76.8 | 88 | 94.7 | 89.5 | 74.2 | 91.7 | 85.8 | |
| CATBackbone=ResNet-502019.03 | 76.7 | 89 | 94.5 | 89.8 | 74 | 93.7 | 86.3 | |
| JAN+CATBackbone=ResNet-502019.03 | 76.3 | 89.2 | 95.3 | 89.3 | 75.9 | 92.2 | 86.4 | |
| RevGradBackbone=ResNet-502019.03 | 75 | 86 | 96.2 | 87 | 74.3 | 91.5 | 85 | |
| ResNet-50Backbone=ResNet-502019.03 | 74.8 | 83.9 | 91.5 | 78 | 65.5 | 91.2 | 80.7 | |
| DANBackbone=ResNet-502019.03 | 74.5 | 82.2 | 92.8 | 86.3 | 69.2 | 89.8 | 82.5 | |
| rRevGrad+CATBackbone=AlexNet2019.03 | 68.6 | 84.6 | 91.9 | 80.8 | 65.6 | 92.5 | 80.7 | |
| CATBackbone=AlexNet2019.03 | 68.3 | 83.6 | 91.3 | 79.1 | 64 | 90.9 | 79.5 | |
| JAN+CATBackbone=AlexNet2019.03 | 67.8 | 83.7 | 92.3 | 80.8 | 65.8 | 91.1 | 80.3 | |
| RTNBackbone=AlexNet2019.03 | 67.4 | 81.3 | 89.5 | 78 | 62 | 89.1 | 77.9 | |
| MSTNBackbone=AlexNet2019.03 | 67.3 | 82.8 | 91.5 | 81.7 | 65.3 | 91.2 | 80 | |
| JANBackbone=AlexNet2019.03 | 67.2 | 82.8 | 91.3 | 80 | 63.5 | 91 | 79.3 | |
| RevGradBackbone=AlexNet2019.03 | 66.5 | 81.8 | 89 | 79.8 | 63.5 | 88.7 | 78.2 | |
| AlexNetBackbone=AlexNet2019.03 | 66.2 | 70 | 84.3 | 71.3 | 59.3 | 84.5 | 73.9 |