Focus on Hiders: Exploring Hidden Threats for Enhancing Adversarial Training

Adversarial training is often formulated as a min-max problem, however, concentrating only on the worst adversarial examples causes alternating repetitive confusion of the model, i.e., previously defended or correctly classified samples are not defensible or accurately classifiable in subsequent adv...

Full description

Saved in:
Bibliographic Details
Published inarXiv.org
Main Authors Li, Qian, Hu, Yuxiao, Dong, Yinpeng, Zhang, Dongxiao, Chen, Yuntian
Format Paper
LanguageEnglish
Published Ithaca Cornell University Library, arXiv.org 12.12.2023
Subjects
Algorithms
Iterative methods
Optimization
Robustness
Online AccessGet full text

Cover

More Information
Summary:Adversarial training is often formulated as a min-max problem, however, concentrating only on the worst adversarial examples causes alternating repetitive confusion of the model, i.e., previously defended or correctly classified samples are not defensible or accurately classifiable in subsequent adversarial training. We characterize such non-ignorable samples as "hiders", which reveal the hidden high-risk regions within the secure area obtained through adversarial training and prevent the model from finding the real worst cases. We demand the model to prevent hiders when defending against adversarial examples for improving accuracy and robustness simultaneously. By rethinking and redefining the min-max optimization problem for adversarial training, we propose a generalized adversarial training algorithm called Hider-Focused Adversarial Training (HFAT). HFAT introduces the iterative evolution optimization strategy to simplify the optimization problem and employs an auxiliary model to reveal hiders, effectively combining the optimization directions of standard adversarial training and prevention hiders. Furthermore, we introduce an adaptive weighting mechanism that facilitates the model in adaptively adjusting its focus between adversarial examples and hiders during different training periods. We demonstrate the effectiveness of our method based on extensive experiments, and ensure that HFAT can provide higher robustness and accuracy.
ISSN:2331-8422