Image Classification on CIFAR10 (val) (Clean Accuracy and ASR)
85.95Clean Accuracyclean data
Evaluation results
| Method | Links | ||
|---|---|---|---|
| clean dataSampler=DDIM, Backbone=ResNet2026.06 | 85.95 | 10.07 | |
| same-classSampler=DDPM, Backbone=ResNet2026.06 | 78.96 | 99.36 | |
| same-classSampler=DDIM, Backbone=ResNet2026.06 | 78.18 | 67.47 | |
| clean dataSampler=DDPM, Backbone=ResNet2026.06 | 77.29 | 10.01 | |
| one-to-oneSampler=DDIM, Backbone=ResNet2026.06 | 76.08 | 84.61 | |
| all-to-oneSampler=DDPM, Backbone=ResNet2026.06 | 75.73 | 97.11 | |
| one-to-oneSampler=DDPM, Backbone=ResNet2026.06 | 75.52 | 95.97 | |
| all-to-oneSampler=DDIM, Backbone=ResNet2026.06 | 74.38 | 81.48 |