English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Camera-based autonomous systems that emulate human perception areincreasingly being integrated into safety-critical platforms. Consequently, anestablished body of literature has emerged that explores adversarial attackstargeting the underlying machine learning models. Adapting adversarial attacksto the physical world is desirable for the attacker, as this removes the needto compromise digital systems. However, the real world poses challenges relatedto the "survivability" of adversarial manipulations given environmental noisein perception pipelines and the dynamicity of autonomous systems. In thispaper, we take a sensor-first approach. We present EvilEye, a man-in-the-middleperception attack that leverages transparent displays to generate dynamicphysical adversarial examples. EvilEye exploits the camera's optics to inducemisclassifications under a variety of illumination conditions. To generatedynamic perturbations, we formalize the projection of a digital attack into thephysical domain by modeling the transformation function of the captured imagethrough the optical pipeline. Our extensive experiments show that EvilEye'sgenerated adversarial perturbations are much more robust across varyingenvironmental light conditions relative to existing physical perturbationframeworks, achieving a high attack success rate (ASR) while bypassingstate-of-the-art physical adversarial detection frameworks. We demonstrate thatthe dynamic nature of EvilEye enables attackers to adapt adversarial examplesacross a variety of objects with a significantly higher ASR compared tostate-of-the-art physical world attack frameworks. Finally, we discussmitigation strategies against the EvilEye attack.

Why Don't You Clean Your Glasses? Perception Attacks with Dynamic Optical Perturbations