Image Classification on MNIST Domain Generalization (Target Domains Average)
62.86Acc (SVHN)L2D
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| L2DBackbone=LeNet, Source Domain=MNIST, Input Resolution=32x32, Batch Size=32, Optimizer=SGD2021.08 | 62.86 | 87.3 | 63.72 | 83.97 | 74.46 | |
| GLOT-DRBackbone=LeNet5, Source Domain=MNIST, Number of runs=10, n=42022.03 | 43.1 | 68.44 | 50.49 | 82.48 | 61.13 | |
| GLOT-DRBackbone=LeNet5, Source Domain=MNIST, Number of runs=10, n=12022.03 | 42.7 | 67.72 | 50.53 | 82.32 | 60.82 | |
| ME-ADABackbone=LeNet, Source Domain=MNIST, Input Resolution=32x32, Batch Size=32, Optimizer=SGD2021.08 | 42.56 | 63.27 | 50.39 | 81.04 | 59.32 | |
| M-ADABackbone=LeNet, Source Domain=MNIST, Input Resolution=32x32, Batch Size=32, Optimizer=SGD2021.08 | 42.55 | 67.94 | 48.95 | 78.53 | 59.49 | |
| GLOT-DRBackbone=LeNet5, Source Domain=MNIST, Number of runs=10, n=22022.03 | 42.35 | 67.95 | 50.53 | 82.33 | 60.81 | |
| ME-ADABackbone=LeNet5, Source Domain=MNIST, Number of runs=102022.03 | 42 | 63.98 | 49.8 | 79.1 | 58.72 | |
| PARBackbone=LeNet5, Source Domain=MNIST, Number of runs=102022.03 | 36.08 | 61.16 | 45.48 | 79.95 | 55.67 | |
| ADABackbone=LeNet5, Source Domain=MNIST, Number of runs=102022.03 | 35.7 | 58.65 | 47.18 | 80.4 | 55.48 | |
| ADABackbone=LeNet, Source Domain=MNIST, Input Resolution=32x32, Batch Size=32, Optimizer=SGD2021.08 | 35.51 | 60.41 | 45.32 | 77.26 | 54.62 | |
| JiGenBackbone=LeNet, Source Domain=MNIST, Input Resolution=32x32, Batch Size=32, Optimizer=SGD2021.08 | 33.8 | 57.8 | 43.79 | 77.15 | 53.14 | |
| Standard (ERM)Backbone=LeNet5, Source Domain=MNIST, Number of runs=102022.03 | 31.95 | 55.96 | 43.85 | 79.92 | 52.92 | |
| ERMBackbone=LeNet, Source Domain=MNIST, Input Resolution=32x32, Batch Size=32, Optimizer=SGD2021.08 | 27.83 | 52.72 | 39.65 | 76.94 | 49.29 | |
| d-SNEBackbone=LeNet, Source Domain=MNIST, Input Resolution=32x32, Batch Size=32, Optimizer=SGD2021.08 | 26.22 | 50.98 | 37.83 | 93.16 | 52.05 | |
| CCSABackbone=LeNet, Source Domain=MNIST, Input Resolution=32x32, Batch Size=32, Optimizer=SGD2021.08 | 25.89 | 49.29 | 37.31 | 83.72 | 49.05 |