LP-GAN: Learning perturbations based on generative adversarial networks for point cloud adversarial attacks

作者：

Highlights：

• We propose a novel adversarial attack method for 3D model, which focuses on the characteristics of the adversarial examples.

• We propose a perception loss based on the spatial similarity, which can improve the quality of the adversarial samples.

• The experimental results with the state-of-the-art methods demonstrate the superiority of our method.

摘要

•We propose a novel adversarial attack method for 3D model, which focuses on the characteristics of the adversarial examples.•We propose a perception loss based on the spatial similarity, which can improve the quality of the adversarial samples.•The experimental results with the state-of-the-art methods demonstrate the superiority of our method.

论文关键词：3D model,Point cloud,Adversarial attack,GAN;

论文评审过程：Received 19 October 2021, Revised 12 December 2021, Accepted 22 December 2021, Available online 29 December 2021, Version of Record 9 February 2022.

论文官网地址：https://doi.org/10.1016/j.imavis.2021.104370