忙完了其它的,还是要回归科研做好整理,相当于论文笔记吧。对抗样本去噪算法算是告以段落,或者说叫对抗样本提纯。本文总结了一些近几年的、思路还行结果也还好的对抗样本去噪算法,就相当于写个类似的综述了,注意,并非详解。优缺点仅是个人分析,其实论文读多了或者看了代码,总会有一些想法。包含以下论文:
- Comdefend: An efficient image compression model to defend adversarial examples, CVPR 2019
- Feature denoising for improving adversarial robustness, CVPR 2019
- Defense against adversarial attacks using high-level representation guided denoiser, CVPR 2018
- A Self-supervised Approach for Adversarial Robustness, CVPR 2020
- Denoised Smoothing: A Provable Defense for Pretrained Classifiers, NIPS 2020
- Stochastic Security: Adversarial Defense Using Long-Run Dynamics of Energy-Based Models, ICLR 2021
- Online Adversarial Purification based on Self-Supervision, ICLR 2021
- Adversarial Purification with Score-based Generative Models, ICML 2021