Generating 3D Adversarial Point Clouds

TL;DR

This paper introduces novel algorithms to generate adversarial point clouds that can deceive 3D deep learning models like PointNet, highlighting vulnerabilities in 3D data processing for safety-critical applications.

Contribution

It proposes new adversarial attack methods for 3D point clouds, including point perturbation and generation, with tailored metrics and high success rates.

Findings

Achieves over 99% success rate in targeted attacks

Demonstrates vulnerability of PointNet to adversarial point clouds

Provides metrics for measuring perturbations in 3D point clouds

Abstract

Deep neural networks are known to be vulnerable to adversarial examples which are carefully crafted instances to cause the models to make wrong predictions. While adversarial examples for 2D images and CNNs have been extensively studied, less attention has been paid to 3D data such as point clouds. Given many safety-critical 3D applications such as autonomous driving, it is important to study how adversarial point clouds could affect current deep 3D models. In this work, we propose several novel algorithms to craft adversarial point clouds against PointNet, a widely used deep neural network for point cloud processing. Our algorithms work in two ways: adversarial point perturbation and adversarial point generation. For point perturbation, we shift existing points negligibly. For point generation, we generate either a set of independent and scattered points or a small number (1-3) of point clusters with meaningful shapes such as balls and airplanes which could be hidden in the human psyche. In addition, we formulate six perturbation measurement metrics tailored to the attacks in point clouds and conduct extensive experiments to evaluate the proposed algorithms on the ModelNet40 3D shape classification dataset. Overall, our attack algorithms achieve a success rate higher than 99% for all targeted attacks