Image Classification on CIFAR-10 all-to-all setting, WaNet attack (test)
94.49AccuracyBefore
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Beforedefense=None2022.02 | 94.49 | 93.47 | |
| Neural Cleansedefense=Neural Cleanse2022.02 | 94.49 | 93.47 | |
| NADdefense=Neural Attention Distillation2022.02 | 93.68 | 3.05 | |
| PBEclean_data_available=true2022.02 | 93.45 | 0.8 | |
| Fine-tuningdefense=Fine-tuning2022.02 | 93.37 | 7.81 | |
| PBEclean_data_available=false2022.02 | 93.24 | 1.02 | |
| ANPdefense=Adversarial Neuron Pruning2022.02 | 93.21 | 0.99 | |
| Fine-pruningdefense=Fine-pruning2022.02 | 89.02 | 92.53 |