3D Point Cloud Completion with Geometric-Aware Adversarial Augmentation

TL;DR

This paper introduces a novel adversarial augmentation method for 3D point cloud completion that improves neural network performance by generating geometric-aware adversarial samples, enhancing robustness and accuracy.

Contribution

The work presents a new adversarial sample generation technique using principal directions, improving 3D point cloud completion and training efficiency over existing methods.

Findings

Training with proposed adversarial samples improves completion accuracy.

Method maintains geometric features and reduces outliers in adversarial samples.

Enhanced robustness of neural networks on ShapeNet dataset.

Abstract

With the popularity of 3D sensors in self-driving and other robotics applications, extensive research has focused on designing novel neural network architectures for accurate 3D point cloud completion. However, unlike in point cloud classification and reconstruction, the role of adversarial samples in3D point cloud completion has seldom been explored. In this work, we show that training with adversarial samples can improve the performance of neural networks on 3D point cloud completion tasks. We propose a novel approach to generate adversarial samples that benefit both the performance of clean and adversarial samples. In contrast to the PGD-k attack, our method generates adversarial samples that keep the geometric features in clean samples and contain few outliers. In particular, we use principal directions to constrain the adversarial perturbations for each input point. The gradient components in the mean direction of principal directions are taken as adversarial perturbations. In addition, we also investigate the effect of using the minimum curvature direction. Besides, we adopt attack strength accumulation and auxiliary Batch Normalization layers method to speed up the training process and alleviate the distribution mismatch between clean and adversarial samples. Experimental results show that training with the adversarial samples crafted by our method effectively enhances the performance of PCN on the ShapeNet dataset.