Physical Adversarial Attacks on an Aerial Imagery Object Detector

TL;DR

This paper explores the vulnerability of aerial imagery object detectors to physical adversarial patches, demonstrating their effectiveness despite atmospheric challenges and highlighting security concerns for satellite image analysis.

Contribution

It presents one of the first studies on physical adversarial attacks on aerial imagery, including optimization, fabrication, and deployment of patches to fool satellite-based object detectors.

Findings

Physical adversarial patches significantly reduce detector accuracy.

Atmospheric factors impact the effectiveness of adversarial patches.

The study highlights security risks in satellite imagery processing.

Abstract

Deep neural networks (DNNs) have become essential for processing the vast amounts of aerial imagery collected using earth-observing satellite platforms. However, DNNs are vulnerable towards adversarial examples, and it is expected that this weakness also plagues DNNs for aerial imagery. In this work, we demonstrate one of the first efforts on physical adversarial attacks on aerial imagery, whereby adversarial patches were optimised, fabricated and installed on or near target objects (cars) to significantly reduce the efficacy of an object detector applied on overhead images. Physical adversarial attacks on aerial images, particularly those captured from satellite platforms, are challenged by atmospheric factors (lighting, weather, seasons) and the distance between the observer and target. To investigate the effects of these challenges, we devised novel experiments and metrics to evaluate the efficacy of physical adversarial attacks against object detectors in aerial scenes. Our results indicate the palpable threat posed by physical adversarial attacks towards DNNs for processing satellite imagery.