Abstract:Deep neural networks (DNNs) have become essential for aerial detection. However, DNNs are vulnerable to adversarial examples, which pose great security concerns for security-critical systems. Researchers recently devised adversarial patches to evaluate the vulnerability of DNN-based aerial detection methods physically. Nonetheless, adversarial patches generated by the existing algorithms are not strong enough and extremely time-consuming. Moreover, complicated physical factors are not accommodated well during the optimization process. In this article, a novel adaptive-patch-based physical attack (AP-PA) framework is proposed to alleviate the above problems, which achieves state-of-the-art performance in both accuracy and efficiency. Specifically, AP-PA aims to generate adversarial patches that are adaptive in both physical dynamics and varying scales, and by which the particular targets can be hidden from being detected. Furthermore, the adversarial patch is also gifted with attack effectiveness against all targets of the same class with a patch outside the target (no need to smear targeted objects) and robust enough in the physical world. In addition, a new loss is devised to consider more available information of detected objects to optimize the adversarial patch, which can significantly improve the patch's attack efficacy (average precision drop up to 87.86% and 85.48% in white-box and black-box settings, respectively) and optimizing efficiency. We also establish one of the first comprehensive, coherent, and rigorous benchmarks to evaluate the attack efficacy of adversarial patches on aerial detection tasks. Finally, several proportionally scaled experiments are performed physically to demonstrate that the elaborated adversarial patches can successfully deceive aerial detection algorithms in dynamic physical circumstances. The code is available at https://github.com/JiaweiLian/AP-PA.

Unified Adversarial Patch for Cross-modal Attacks in the Physical World

Unified Adversarial Patch for Visible-Infrared Cross-modal Attacks in the Physical World

Two-stage optimized unified adversarial patch for attacking visible-infrared cross-modal detectors in the physical world

Physically Adversarial Infrared Patches with Learnable Shapes and Locations

Towards a Physical-World Adversarial Patch for Blinding Object Detection Models.

Contextual Adversarial Attack Against Aerial Detection in the Physical World

Universal Physical Camouflage Attacks on Object Detectors

Learning Coated Adversarial Camouflages for Object Detectors

TPatch: A Triggered Physical Adversarial Patch

Infrared Adversarial Patch Generation Based on Reinforcement Learning

Universal Physical Adversarial Attack via Background Image

DePatch: Towards Robust Adversarial Patch for Evading Person Detectors in the Real World

Adversarial Camera Patch: An Effective and Robust Physical-World Attack on Object Detectors

Camouflaged Adversarial Patch Attack on Object Detector

Adversarial Patch Attack on Multi-Scale Object Detection for UAV Remote Sensing Images

Physical Attack on Monocular Depth Estimation with Optimal Adversarial Patches

Benchmarking Adversarial Patch Against Aerial Detection

Physical Adversarial Attack Scheme on Object Detectors using 3D Adversarial Object

Defending Against Physical Adversarial Patch Attacks on Infrared Human Detection

Towards Physically-Realizable Adversarial Attacks in Embodied Vision Navigation

Empirical Evaluation of Physical Adversarial Patch Attacks Against Overhead Object Detection Models