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

Deep neural networks (DNNs) are inherently vulnerable to adversarial inputs:such maliciously crafted samples trigger DNNs to misbehave, leading todetrimental consequences for DNN-powered systems. The fundamental challenges ofmitigating adversarial inputs stem from their adaptive and variable nature.Existing solutions attempt to improve DNN resilience against specific attacks;yet, such static defenses can often be circumvented by adaptively engineeredinputs or by new attack variants.  Here, we present EagleEye, an attack-agnostic adversarial tampering analysisengine for DNN-powered systems. Our design exploits the {\em minimalityprinciple} underlying many attacks: to maximize the attack's evasiveness, theadversary often seeks the minimum possible distortion to convert genuine inputsto adversarial ones. We show that this practice entails the distinctdistributional properties of adversarial inputs in the input space. Byleveraging such properties in a principled manner, EagleEye effectivelydiscriminates adversarial inputs and even uncovers their correct classificationoutputs. Through extensive empirical evaluation using a range of benchmarkdatasets and DNN models, we validate EagleEye's efficacy. We furtherinvestigate the adversary's possible countermeasures, which implies a difficultdilemma for her: to evade EagleEye's detection, excessive distortion isnecessary, thereby significantly reducing the attack's evasiveness regardingother detection mechanisms.

EagleEye: Attack-Agnostic Defense against Adversarial Inputs (Technical Report)