Boosting the Transferability of Adversarial Attacks in Deep Neural Networks
Cambridge Explorations in Arts and Sciences · 2023
DOI: 10.61603/ceas.v1i2.19 ↗Kurz gefasst
Fachartikel in der Zeitschrift Cambridge Explorations in Arts and Sciences, 2023. Verfasst von Xiaotang Xu und 4 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-SA 4.0).
Abstract (Originalsprache)
This article presents a means of boosting the transferability of adversarial attacks in deep neural networks. The research includes background, methodologies, and outcomes, encompassing single-attack approaches and ensemble attack strategies such as I-FGSM and MI-FGSM. We delve into the notion of retraining using adversarial examples. Our contributions reveal the limitations of single-attack methods regarding transferability and demonstrate the superiority of ensemble attack methods. We highlight how algorithm selection impacts attack effectiveness and how model variations enhance transferability. Through these investigations, we offer valuable insights for bolstering deep neural networks’ adversarial robustness while acknowledging existing constraints.
Zitieren
APA
Xu, X., Wei, K., Xu, J., Zhu, F., & Zhang, J. (2023). Boosting the Transferability of Adversarial Attacks in Deep Neural Networks. Cambridge Explorations in Arts and Sciences. https://doi.org/10.61603/ceas.v1i2.19
Vancouver
Xu X, Wei K, Xu J, Zhu F, Zhang J. Boosting the Transferability of Adversarial Attacks in Deep Neural Networks. Cambridge Explorations in Arts and Sciences. 2023. doi:10.61603/ceas.v1i2.19
BibTeX
@article{Xu2023,
author = {Xu, Xiaotang and Wei, Kangxin and Xu, Jin and Zhu, Fangyi and Zhang, Jiayuan},
title = {Boosting the Transferability of Adversarial Attacks in Deep Neural Networks},
journal = {Cambridge Explorations in Arts and Sciences},
year = {2023},
doi = {10.61603/ceas.v1i2.19}
}
EUVIMED