Unsupervised Image To Image Translation On
評価指標
Classification Accuracy
評価結果
このベンチマークにおける各モデルのパフォーマンス結果
モデル名 | Classification Accuracy | Paper Title | Repository |
|---|---|---|---|
| CyCADA pixel+feat | 90.4% | CyCADA: Cycle-Consistent Adversarial Domain Adaptation | - |
| DANN | 73.6% | Unsupervised Domain Adaptation by Backpropagation | - |
| ADDA | 76.0% | Adversarial Discriminative Domain Adaptation | - |
| DTN | 84.4% | Unsupervised Cross-Domain Image Generation | - |
0 of 4 row(s) selected.