- Rating
- 2
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 1
Read review
Paper summary
This job is mainly dedicated to solving the problem of slow inference speed in adversarial purification processes based on generative models. The author provided experimental results on ImageNet-1K and CIFAR-10, and theoretical analysis validated the relevant claims.
Critical strengths
- The main experiments in this article are well conducted, verifying the robustness of ResNet-152 and WideResNet-50-2.
- The effectiveness of the method was further qualitatively analyzed through interpretability experiments.
Critical weaknesses
- This article only conducted experiments on CNN, not on ViT.
- This article only includes BPDA-PGD experiments, not BPDA-EOT experiments.
- The attack cost and number of steps in this article are too small, especially the number of attack steps.
- The AA results should be compared on a standard benchmark.
- No comparison was made with AT.
- The experiment can be implemented on both and -norm attack settings.
- Table 6 shows that the purification strength hyperparameter is highly sensitive, with inherent trade-offs and no adaptive mechanism. In recent years, there have been multiple purification works. This paper proposes a plugin, but it should demonstrate its universality rather than simply adjusting parameters to obtain results.
Questions for authors
- Could the experimetns on ViT be supplemented on the maanuscript?
- How could you explain the significant attenuation of robustness under strong adaptive white box attacks?
- Why does this article avoid the gradient masking problem and dare not directly compare with AT?
- What does the experimental result in Figure 4 want to prove?
- Inference heavily relies on Gaussian noise sampling, introducing random factors. Why did you only conduct one experiment instead of multiple sub-experiments?