Generating 3D Adversarial 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

arXiv commentCVPR 2019

References in corpus (15)

• PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation
• PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space
• Feature Squeezing: Detecting Adversarial Examples in Deep Neural Networks
• Synthesizing Robust Adversarial Examples
• Robust Physical-World Attacks on Deep Learning Models
• Generating Natural Adversarial Examples
• Dynamic Graph CNN for Learning on Point Clouds
• Defense-GAN: Protecting Classifiers Against Adversarial Attacks Using Generative Models
• Adversarial Examples for Evaluating Reading Comprehension Systems
• Spatially Transformed Adversarial Examples
• Defensive Distillation is Not Robust to Adversarial Examples
• Generating Adversarial Examples with Adversarial Networks
• Deep Defense: Training DNNs with Improved Adversarial Robustness
• Improving Network Robustness against Adversarial Attacks with Compact Convolution
• A Deeper Look at 3D Shape Classifiers

Cited by in corpus (9)

• Adversarial Attack and Defense on Point Sets
• 3D-VField: Adversarial Augmentation of Point Clouds for Domain Generalization in 3D Object Detection
• Adversarial Objects Against LiDAR-Based Autonomous Driving Systems
• Extending Adversarial Attacks and Defenses to Deep 3D Point Cloud Classifiers
• Towards Understanding the Adversarial Vulnerability of Skeleton-based Action Recognition
• PointCloud Saliency Maps
• Rearchitecting Classification Frameworks For Increased Robustness
• Adversarial shape perturbations on 3D point clouds
• Generating Band-Limited Adversarial Surfaces Using Neural Networks