Boosting 3D Adversarial Attacks With Attacking on Frequency

Binbin Liu, Jinlai Zhang, Jihong Zhu

Abstract: Deep neural networks (DNNs) have been shown to be vulnerable to adversarial attacks in the image domain. Recently, 3D adversarial attacks, especially adversarial attacks on point clouds, have elicited mounting interest. However, adversarial point clouds obtained by previous methods show weak transferability and are easy to defend. To address these problems, in this paper we propose a strong point cloud attack method named AOF which pays more attention to the low-frequency component of point clouds. We combine the losses from point cloud and its low-frequency component to craft adversarial samples and focus on the low-frequency component of point cloud in the process of optimization. Extensive experiments validate that AOF can improve the transferability significantly compared to state-of-the-art (SOTA) attacks, and is more robust to state-of-the-art 3D defense methods. Otherwise, compared to adversarial point clouds generated by other adversarial attack methods, adversarial point clouds obtained by AOF contain more deformation than outlier. Code is available at: https://github.com/code-roamer/AOF .