Unsupervised Domain Adaptation
Unsupervised Domain Adaptation is a learning framework to transfer knowledge learned from source domains with a large number of annotated training examples to target domains with unlabeled data only.
Source: Domain-Specific Batch Normalization for Unsupervised Domain Adaptation
Papers
Showing 1–10 of 1951 papers
All datasetsDuke to MarketMarket to DukeCityscapes-to-Foggy CityscapesOffice-HomeMarket to MSMTImageNet-CSYNTHIA-to-CityscapesVehicleID to VeRi-776Duke to MSMTSIM10K to CityscapesVeri-776 to VehicleID LargeVeri-776 to VehicleID Medium
Benchmark Results
| # | Model | Metric | Claimed | Verified | Status |
|---|---|---|---|---|---|
| 1 | ResNet50 (baseline), BatchNorm Adaptation, 8 samples | mean Corruption Error (mCE) | 65 | — | Unverified |
| 2 | ResNet50 (baseline), BatchNorm Adaptation, full adaptation | mean Corruption Error (mCE) | 62.2 | — | Unverified |
| 3 | ResNet50 + ENT | mean Corruption Error (mCE) | 51.6 | — | Unverified |
| 4 | ResNet50 + RPL | mean Corruption Error (mCE) | 50.5 | — | Unverified |
| 5 | ResNet50+DeepAug+AugMix, BatchNorm Adaptation, 8 samples | mean Corruption Error (mCE) | 48.4 | — | Unverified |
| 6 | ResNet50+DeepAug+AugMix, BatchNorm Adaptation, full adaptation | mean Corruption Error (mCE) | 45.4 | — | Unverified |
| 7 | ResNeXt101 32x8d + ENT | mean Corruption Error (mCE) | 44.3 | — | Unverified |
| 8 | ResNeXt101 32x8d + RPL | mean Corruption Error (mCE) | 43.2 | — | Unverified |
| 9 | ResNeXt101 32x8d + IG-3.5B + RPL | mean Corruption Error (mCE) | 40.9 | — | Unverified |
| 10 | ResNeXt101 32x8d + IG-3.5B + ENT | mean Corruption Error (mCE) | 40.8 | — | Unverified |