Rethinking Multi-domain Generalization with A General Learning Objective

Open paper
Benchmark Model Rank Results
domain-generalization-on-domainnetGMDG (RegNetY-16GF, SWAD)#7Average Accuracy: 61.3
domain-generalization-on-domainnetGMDG (RegNetY-16GF)#16Average Accuracy: 54.6
domain-generalization-on-domainnetGMDG (ResNet-50, SWAD)#23Average Accuracy: 47.3
domain-generalization-on-domainnetGMDG (ResNet-50)#28Average Accuracy: 44.6
domain-generalization-on-office-homeGMDG (RegNetY-16GF, SWAD)#6Average Accuracy: 84.7
domain-generalization-on-office-homeGMDG (RegNetY-16GF)#13Average Accuracy: 80.8
domain-generalization-on-office-homeGMDG (ResNet-50, SWAD)#20Average Accuracy: 72.5
domain-generalization-on-office-homeGMDG (ResNet-50)#25Average Accuracy: 70.7
domain-generalization-on-pacs-2GMDG (RegNetY-16GF, SWAD)#3Average Accuracy: 97.9
domain-generalization-on-pacs-2GMDG (e RegNetY-16GF)#6Average Accuracy: 97.3
domain-generalization-on-pacs-2GMDG (ResNet-50, SWAD)#26Average Accuracy: 88.4
domain-generalization-on-pacs-2GMDG (ResNet-50)#43Average Accuracy: 85.6
domain-generalization-on-terraincognitaGMDG (RegNetY-16GF, SWAD)#2Average Accuracy: 65
domain-generalization-on-terraincognitaGMDG (RegNetY-16GF)#5Average Accuracy: 60.7
domain-generalization-on-terraincognitaGMDG (ResNet-50, SWAD)#11Average Accuracy: 53.0
domain-generalization-on-terraincognitaGMDG (ResNet-50)#17Average Accuracy: 51.1
domain-generalization-on-vlcsGMDG (RegNetY-16GF)#9Average Accuracy: 82.4
domain-generalization-on-vlcsGMDG (RegNetY-16GF, SWAD)#12Average Accuracy: 82.2
domain-generalization-on-vlcsGMDG (ResNet-50, SWAD)#21Average Accuracy: 79.6
domain-generalization-on-vlcsGMDG (ResNet-50)#24Average Accuracy: 79.2

Rank counts only results with a code link.