Adversarial 3D Virtual Patches using Integrated Gradients

TL;DR

This paper introduces a novel virtual patch-based attack method on LiDAR sensors that significantly reduces the spoofing area needed to hide objects, utilizing Integrated Gradients for identifying critical regions.

Contribution

It presents a new object-hiding strategy using virtual patches and a framework to identify critical LiDAR regions, reducing spoofing area while maintaining attack success.

Findings

VP-focused attacks achieve similar success with less spoofing area

Saliency-LiDAR identifies critical regions effectively

CVPs reduce detection recall by at least 15%

Abstract

LiDAR sensors are widely used in autonomous vehicles to better perceive the environment. However, prior works have shown that LiDAR signals can be spoofed to hide real objects from 3D object detectors. This study explores the feasibility of reducing the required spoofing area through a novel object-hiding strategy based on virtual patches (VPs). We first manually design VPs (MVPs) and show that VP-focused attacks can achieve similar success rates with prior work but with a fraction of the required spoofing area. Then we design a framework Saliency-LiDAR (SALL), which can identify critical regions for LiDAR objects using Integrated Gradients. VPs crafted on critical regions (CVPs) reduce object detection recall by at least 15% compared to our baseline with an approximate 50% reduction in the spoofing area for vehicles of average size.