Adversarial Patch for 3D Local Feature Extractor

TL;DR

This paper explores adversarial patch attacks on 3D local feature extractors, demonstrating how to manipulate feature matching to deceive computer vision systems, and discusses various patch generation techniques.

Contribution

It introduces methods to generate adversarial patches targeting 3D local feature extractors, revealing vulnerabilities and potential defenses in feature matching processes.

Findings

Adversarial patches can force non-matching regions to match

Patches can prevent matching of originally matching regions

Different patch generation methods have varying effectiveness and drawbacks

Abstract

Local feature extractors are the cornerstone of many computer vision tasks. However, their vulnerability to adversarial attacks can significantly compromise their effectiveness. This paper discusses approaches to attack sophisticated local feature extraction algorithms and models to achieve two distinct goals: (1) forcing a match between originally non-matching image regions, and (2) preventing a match between originally matching regions. At the end of the paper, we discuss the performance and drawbacks of different patch generation methods.