Image Classification on Visda 2021
64.8AccuracyPCSR
Evaluation results
| Method | Links | |
|---|---|---|
| PCSRadaptation_mode=test-time adaptation2025.12 | 64.8 | |
| TENTadaptation_mode=test-time adaptation2025.12 | 60.1 | |
| SARadaptation_mode=test-time adaptation2025.12 | 60.1 | |
| DeYOBackbone=VitBase-LN, Batch Size=1, Single Sample Adaptation=true2024.03 | 58.7 | |
| Sourceadaptation_mode=test-time adaptation2025.12 | 57.7 | |
| SARBackbone=VitBase-LN, Batch Size=1, Single Sample Adaptation=true2024.03 | 51.2 | |
| TentBackbone=VitBase-LN, Batch Size=1, Single Sample Adaptation=true2024.03 | 50.4 | |
| Gradadaptation_mode=test-time2026.06 | 48.46 | |
| EATABackbone=VitBase-LN, Batch Size=1, Single Sample Adaptation=true2024.03 | 47.9 | |
| SVGDadaptation_mode=test-time2026.06 | 47.11 | |
| KLadaptation_mode=test-time2026.06 | 46.91 | |
| DeYOBackbone=ResNet-50-GN, Batch Size=1, Single Sample Adaptation=true2024.03 | 45.9 | |
| Naiveadaptation_mode=test-time2026.06 | 45.81 | |
| DeYoadaptation_mode=test-time2026.06 | 45.41 | |
| No Adapt.Backbone=VitBase-LN, Batch Size=1, Single Sample Adaptation=true2024.03 | 44.3 | |
| SARadaptation_mode=test-time2026.06 | 43.94 | |
| TentBackbone=ResNet-50-GN, Batch Size=1, Single Sample Adaptation=true2024.03 | 43.9 | |
| EATABackbone=ResNet-50-GN, Batch Size=1, Single Sample Adaptation=true2024.03 | 43.9 | |
| Tentadaptation_mode=test-time2026.06 | 43.88 | |
| SARBackbone=ResNet-50-GN, Batch Size=1, Single Sample Adaptation=true2024.03 | 43.8 | |
| No Adapt.Backbone=ResNet-50-GN, Batch Size=1, Single Sample Adaptation=true2024.03 | 43.5 |