Single-source Domain Generalization on PACS (test)
72.2Average Domain Transfer AccuracyMEDIC++
Evaluation results
| Method | Links | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| MEDIC++Backbone=ResNet-502026.06 | 72.2 | — | — | — | — | — | — | — | — | — | — | — | — | 60.6 | 82 | 75.4 | 70.8 | |
| START-MBackbone=VMamba, Evaluation protocol=Single-source domain generalization (SDG)2024.10 | 71.64 | — | — | — | — | — | — | — | — | — | — | — | — | 45.02 | 85.44 | 78.08 | 78.03 | |
| START-XBackbone=VMamba, Evaluation protocol=Single-source domain generalization (SDG)2024.10 | 71.54 | — | — | — | — | — | — | — | — | — | — | — | — | 44.87 | 85.66 | 79.4 | 76.24 | |
| MEDICBackbone=ResNet-502026.06 | 69.9 | — | — | — | — | — | — | — | — | — | — | — | — | 54.3 | 81.1 | 74.7 | 69.6 | |
| Strong baselineBackbone=VMamba, Evaluation protocol=Single-source domain generalization (SDG)2024.10 | 69.52 | — | — | — | — | — | — | — | — | — | — | — | — | 44.92 | 83.29 | 75.1 | 74.76 | |
| ABAconvolutional layers=3-layers2023.07 | 66.36 | — | — | — | — | — | — | — | — | — | — | — | — | 58.86 | 77.49 | 75.34 | 53.76 | |
| ABAconvolutional layers=5-layers2023.07 | 66.02 | — | — | — | — | — | — | — | — | — | — | — | — | 59.04 | 77.16 | 74.71 | 53.18 | |
| ABAconvolutional layers=5-layers, additional module=RandConv2023.07 | 66 | — | — | — | — | — | — | — | — | — | — | — | — | 57.59 | 76.66 | 75.61 | 54.12 | |
| ABAconvolutional layers=3-layers, additional module=RandConv2023.07 | 65.76 | — | — | — | — | — | — | — | — | — | — | — | — | 56.95 | 77.21 | 75.34 | 53.52 | |
| ABAconvolutional layers=5-layers, additional module=Augmix2023.07 | 65.55 | — | — | — | — | — | — | — | — | — | — | — | — | 57.87 | 77.29 | 74.7 | 52.35 | |
| ALTconvolutional layers=5-layers, additional module=Augmix2023.07 | 64.72 | — | — | — | — | — | — | — | — | — | — | — | — | 55.09 | 77.36 | 75.69 | 50.72 | |
| ABAconvolutional layers=1-layer2023.07 | 64.63 | — | — | — | — | — | — | — | — | — | — | — | — | 54.49 | 75.61 | 75.59 | 52.84 | |
| ALTconvolutional layers=5-layers, additional module=RandConv2023.07 | 64.18 | — | — | — | — | — | — | — | — | — | — | — | — | 55.66 | 76.23 | 73.96 | 50.86 | |
| ALTconvolutional layers=5-layers2023.07 | 63.6 | — | — | — | — | — | — | — | — | — | — | — | — | 54.33 | 75.96 | 74.06 | 50.03 | |
| ABAconvolutional layers=1-layer, additional module=RandConv2023.07 | 63.58 | — | — | — | — | — | — | — | — | — | — | — | — | 52.32 | 76.01 | 75.77 | 50.2 | |
| ALTconvolutional layers=1-layer, additional module=RandConv2023.07 | 62.51 | — | — | — | — | — | — | — | — | — | — | — | — | 52.24 | 75.16 | 73.46 | 49.21 | |
| CMBackbone=ResNet-502026.06 | 62.1 | — | — | — | — | — | — | — | — | — | — | — | — | 47 | 75.1 | 67.5 | 58.7 | |
| SagNetBackbone=ResNet-182019.10 | 61.9 | 67.1 | 56.8 | 95.7 | 72.1 | 69.2 | 85.7 | 41.1 | 62.9 | 46.2 | 69.8 | 35.1 | 40.7 | — | — | — | — | |
| ALTconvolutional layers=1-layer2023.07 | 61.88 | — | — | — | — | — | — | — | — | — | — | — | — | 50.83 | 75 | 73.87 | 47.83 | |
| SagNet2023.07 | 61.86 | — | — | — | — | — | — | — | — | — | — | — | — | 48.53 | 75.66 | 73.2 | 50.06 | |
| Augmix2023.07 | 60.51 | — | — | — | — | — | — | — | — | — | — | — | — | 45.24 | 74.66 | 71.47 | 47.72 | |
| RandConvconvolutional layers=1-layer2023.07 | 60.34 | — | — | — | — | — | — | — | — | — | — | — | — | 49.8 | 67.9 | 69.63 | 54.06 | |
| ERMBackbone=ResNet-502026.06 | 59 | — | — | — | — | — | — | — | — | — | — | — | — | 45.6 | 73.4 | 68.5 | 48.7 | |
| ADABackbone=ResNet-182019.10 | 58.7 | 64.3 | 58.5 | 94.5 | 66.7 | 65.6 | 83.6 | 37 | 58.6 | 41.6 | 65.3 | 32.7 | 35.9 | — | — | — | — | |
| ADA2023.07 | 58.68 | — | — | — | — | — | — | — | — | — | — | — | — | 44.63 | 71.96 | 72.43 | 45.73 | |
| JiGen2023.07 | 54.61 | — | — | — | — | — | — | — | — | — | — | — | — | 41.7 | 72.23 | 67.7 | 36.83 | |
| JiGenBackbone=ResNet-182019.10 | 54.6 | 57 | 50 | 96.1 | 65.3 | 65.9 | 85.5 | 26.6 | 41.1 | 42.8 | 62.4 | 27.2 | 35.5 | — | — | — | — | |
| ResNet-18Backbone=ResNet-182019.10 | 54.3 | 62.3 | 49 | 95.2 | 65.7 | 60.7 | 83.6 | 28 | 54.5 | 35.6 | 64.1 | 23.6 | 29.1 | — | — | — | — | |
| ERM2023.07 | 54.28 | — | — | — | — | — | — | — | — | — | — | — | — | 38.93 | 70 | 68.83 | 39.36 | |
| AdvBNN2023.07 | 51.92 | — | — | — | — | — | — | — | — | — | — | — | — | 45.93 | 60.24 | 75.33 | 26.19 |