Towards Transferable Targeted Attack
Authors:
- 1 June 2020
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 638-646
Abstract
An intriguing property of adversarial examples is their transferability, which suggests that black-box attacks are feasible in real-world applications. Previous works mostly study the transferability on non-targeted setting. However, recent studies show that targeted adversarial examples are more difficult to transfer than non-targeted ones. In this paper, we find there exist two defects that lead to the difficulty in generating transferable examples. First, the magnitude of gradient is decreasing during iterative attack, causing excessive consistency between two successive noises in accumulation of momentum, which is termed as noise curing. Second, it is not enough for targeted adversarial examples to just get close to target class without moving away from true class. To overcome the above problems, we propose a novel targeted attack approach to effectively generate more transferable adversarial examples. Specifically, we first introduce the Poincar\'{e} distance as the similarity metric to make the magnitude of gradient self-adaptive during iterative attack to alleviate noise curing. Furthermore, we regularize the targeted attack process with metric learning to take adversarial examples away from true label and gain more transferable targeted adversarial examples. Experiments on ImageNet validate the superiority of our approach achieving 8\% higher attack success rate over other state-of-the-art methods on average in black-box targeted attack.Keywords
This publication has 10 references indexed in Scilit:
- Evading Defenses to Transferable Adversarial Examples by Translation-Invariant Attacks2019
- Improving Transferability of Adversarial Examples With Input Diversity2019
- Adversarial Examples: Attacks and Defenses for Deep LearningIEEE Transactions on Neural Networks and Learning Systems, 2019
- Boosting Adversarial Attacks with Momentum2018
- Robust Physical-World Attacks on Deep Learning Visual Classification2018
- Adversarial Examples Are Not Easily Detected2017
- Towards Evaluating the Robustness of Neural Networks2017
- Rethinking the Inception Architecture for Computer Vision2016
- Deep Metric Learning Using Triplet Network2015
- FaceNet: A unified embedding for face recognition and clustering2015