Domain Generalization on OfficeHome DomainBed (OOD)
85.74Avg OOD AccuracyVL2V-ADIP
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| VL2V-ADIPStudent Backbone=ViT-B/16, Initialization=ImageNet-1K, Strategy=SWAD2023.10 | 85.74 | — | — | — | — | |
| DARTStudent Backbone=ViT-B/16, Initialization=ImageNet-1K2023.10 | 83.75 | — | — | — | — | |
| RISEStudent Backbone=ViT-B/16, Initialization=ImageNet-1K2023.10 | 83.48 | — | — | — | — | |
| ERM Full Fine-tuningStudent Backbone=ViT-B/16, Initialization=ImageNet-1K, Strategy=SWAD2023.10 | 83.22 | — | — | — | — | |
| KDStudent Backbone=ViT-B/16, Initialization=ImageNet-1K, Strategy=SWAD2023.10 | 82.73 | — | — | — | — | |
| SAGMStudent Backbone=ViT-B/16, Initialization=ImageNet-1K2023.10 | 82.22 | — | — | — | — | |
| LP-FTStudent Backbone=ViT-B/16, Initialization=ImageNet-1K2023.10 | 81.55 | — | — | — | — | |
| VL2V-ADIPStudent Backbone=ViT-S/16, Strategy=SWAD2023.10 | 81.22 | — | — | — | — | |
| MIROStudent Backbone=ViT-B/16, Initialization=ImageNet-1K2023.10 | 80.09 | — | — | — | — | |
| VL2V-ADIPStudent Backbone=DeiT-S/16, Strategy=SWAD2023.10 | 77.61 | — | — | — | — | |
| VL2V-ADIPStudent Backbone=ResNet-50, Strategy=SWAD2023.10 | 74.42 | — | — | — | — | |
| ERM-LPStudent Backbone=ViT-B/16, Initialization=ImageNet-1K, Mode=Linear Probing2023.10 | 71.48 | — | — | — | — | |
| Sparse and Shared Feature ActivationsFramework=DomainBed2023.04 | 70.9 | 56.3 | 66.7 | 79.2 | 81.3 | |
| Text2ConceptStudent Backbone=ViT-B/16, Initialization=ImageNet-1K2023.10 | 70.24 | — | — | — | — | |
| ERMFramework=DomainBed2023.04 | 67.5 | 53.4 | 62.7 | 76.5 | 77.3 | |
| SimKDStudent Backbone=ViT-B/16, Initialization=ImageNet-1K2023.10 | 66.76 | — | — | — | — |