3D-VFD: A Victim-free Detector against 3D Adversarial Point Clouds

TL;DR

This paper introduces 3D-VFD, a novel victim-free detector that identifies 3D adversarial point clouds by analyzing geometric feature discrepancies without relying on the victim model's outputs.

Contribution

The paper presents a new 3D steganalysis approach that effectively detects adversarial point clouds without dependence on the target model, outperforming existing methods.

Findings

Achieves state-of-the-art detection accuracy.

Effectively detects point adding and perturbation attacks.

Maintains fast detection speed.

Abstract

3D deep models consuming point clouds have achieved sound application effects in computer vision. However, recent studies have shown they are vulnerable to 3D adversarial point clouds. In this paper, we regard these malicious point clouds as 3D steganography examples and present a new perspective, 3D steganalysis, to counter such examples. Specifically, we propose 3D-VFD, a victim-free detector against 3D adversarial point clouds. Its core idea is to capture the discrepancies between residual geometric feature distributions of benign point clouds and adversarial point clouds and map these point clouds to a lower dimensional space where we can efficiently distinguish them. Unlike existing detection techniques against 3D adversarial point clouds, 3D-VFD does not rely on the victim 3D deep model's outputs for discrimination. Extensive experiments demonstrate that 3D-VFD achieves state-of-the-art detection and can effectively detect 3D adversarial attacks based on point adding and point perturbation while keeping fast detection speed.