Image Classification on Digits MNIST
99.47Average AccuracyOurs
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| OursNumber of Parameters=1,028,2342022.07 | 99.47 | — | — | — | — | |
| Inception-V3Number of Parameters=23,851,7842022.07 | 93.31 | — | — | — | — | |
| VGG-16Number of Parameters=138,357,5442022.07 | 92.4 | — | — | — | — | |
| ME-ADA + AdvStyleBase Method=ME-ADA, Augmentation=AdvStyle2022.07 | 67.3 | 55.5 | 74.1 | 59.3 | 80.1 | |
| ERM + AdvStyleBase Method=ERM, Augmentation=AdvStyle2022.07 | 66 | 50.4 | 73.4 | 58.7 | 81.6 | |
| M-ADA2022.07 | 59.5 | 42.6 | 67.9 | 49 | 78.5 | |
| ME-ADA2022.07 | 59.3 | 42.6 | 63.3 | 50.4 | 81 | |
| ADA2022.07 | 54.6 | 35.5 | 60.4 | 45.3 | 77.3 | |
| JiGen2022.07 | 53.1 | 33.8 | 57.8 | 43.8 | 77.2 | |
| d-SNE2022.07 | 52.1 | 26.2 | 51 | 37.8 | 93.2 | |
| ERM2022.07 | 49.3 | 27.8 | 52.7 | 39.7 | 76.9 | |
| CCSA2022.07 | 49.1 | 25.9 | 49.3 | 37.3 | 83.7 |