Multiple adversarial domains adaptation approach for mitigating adversarial attacks effects

Rasheed, Bader; Khan, Adil; Ahmad, Muhammad; Mazzara, Manuel; Kazmi, S. M.Ahsan

Abstract

Although neural networks are near achieving performance similar to humans in many tasks, they are susceptible to adversarial attacks in the form of a small, intentionally designed perturbation, which could lead to misclassifications. The best defense against these attacks, so far, is adversarial training (AT), which improves a model's robustness by augmenting the training data with adversarial examples. However, AT usually decreases the model's accuracy on clean samples and could overfit to a specific attack, inhibiting its ability to generalize to new attacks. In this paper, we investigate the usage of domain adaptation to enhance AT's performance. We propose a novel multiple adversarial domain adaptation (MADA) method, which looks at this problem as a domain adaptation task to discover robust features. Specifically, we use adversarial learning to learn features that are domain-invariant between multiple adversarial domains and the clean domain. We evaluated MADA on MNIST and CIFAR-10 datasets with multiple adversarial attacks during training and testing. The results of our experiments show that MADA is superior to AT on adversarial samples by about 4% on average and on clean samples by about 1% on average.

• Journal Article Type: Article
• Acceptance Date: Sep 19, 2022
• Publication Date: Oct 10, 2022
• Deposit Date: Nov 9, 2022
• Publicly Available Date: Nov 9, 2022
• Journal: International Transactions on Electrical Energy Systems
• Electronic ISSN: 2050-7038
• Publisher: Wiley
• Peer Reviewed: Peer Reviewed
• Volume: 2022
• Article Number: 2890761
• Pages: 1-11
• DOI: https://doi.org/10.1155/2022/2890761
• Keywords: Electrical and Electronic Engineering, Energy Engineering and Power Technology, Modeling and Simulation
• Public URL: https://uwe-repository.worktribe.com/output/10104175
• Publisher URL: https://www.hindawi.com/journals/itees/2022/2890761/
• Additional Information: Previously reported datasets were used to support this study; and are available at DOI: 10.1109/MSP.2012.2211477 and; DOI: 10.1.1.222.9220. 'ese prior studies and datasets are; cited at relevant places within the text as references [28, 29]