AB-FGSM: AdaBelief optimizer and FGSM-based approach to generate adversarial examples. (August 2022)

• Record Type: Journal Article
• Authors: Wang, Yixiang; Liu, Jiqiang; Chang, Xiaolin; Wang, Jianhua; Rodríguez, Ricardo J.
• Is Part Of: Journal of information security and applications. Volume 68(2022)
• Publication Date: 2022-08
• Subjects: Adversarial Examples -- Deep Learning -- Generalization -- Optimization -- Security -- Transferability; Computer security -- Periodicals; Information technology -- Security measures -- Periodicals; 005.805
• DOI: 10.1016/j.jisa.2022.103227
• Languages: English
• ISSNs: 2214-2126

Abstract

Abstract: Deep neural networks (DNNs) can be misclassified by adversarial examples, which are legitimate inputs integrated with imperceptible perturbations at the testing stage. Extensive research has made progress for white-box adversarial attacks to craft adversarial examples with a high success rate. However, these crafted examples have a low success rate in misleading black-box models with defensive mechanisms. To tackle this problem, we design an A da B elief based iterative F ast G radient S ign M ethod (AB-FGSM) to generalize adversarial examples. By integrating the AdaBelief optimizer into the iterative-FGSM (I-FGSM), the generalization of adversarial examples is boosted, considering that the AdaBelief method can find the transferable adversarial point in the ε ball around the legitimate input on different optimization surfaces. We carry out white-box and black-box attacks on various adversarially trained models and ensemble models to verify the effectiveness and transferability of the adversarial examples crafted by AB-FGSM. Our experimental results indicate that the proposed AB-FGSM can efficiently and effectively craft adversarial examples in the white-box setting compared with state-of-the-art attacks. In addition, the transfer rate of adversarial examples is 4% to 21% higher than that of state-of-the-art attacks in the black-box manner.