Beating White-Box Defenses with Black-Box Attacks

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F21%3A10437329" target="_blank" >RIV/00216208:11320/21:10437329 - isvavai.cz</a>

Výsledek na webu

<a href="https://doi.org/10.1109/IJCNN52387.2021.9533772" target="_blank" >https://doi.org/10.1109/IJCNN52387.2021.9533772</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/IJCNN52387.2021.9533772" target="_blank" >10.1109/IJCNN52387.2021.9533772</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Beating White-Box Defenses with Black-Box Attacks

Popis výsledku v původním jazyce

Deep learning has achieved great results in the last decade, however, it is sensitive to so called adversarial attacks small perturbations of the input that cause the network to classify incorrectly. In the last years a number of attacks and defenses against these attacks were described. Most of the defenses however focus on defending against gradient-based attacks. In this paper, we describe an evolutionary attack and show that the adversarial examples produced by the attack have different features than those from gradient-based attacks. We also show that these features mean that one of the state-of-the-art defenses fails to detect such attacks.

Název v anglickém jazyce

Beating White-Box Defenses with Black-Box Attacks

Popis výsledku anglicky

Deep learning has achieved great results in the last decade, however, it is sensitive to so called adversarial attacks small perturbations of the input that cause the network to classify incorrectly. In the last years a number of attacks and defenses against these attacks were described. Most of the defenses however focus on defending against gradient-based attacks. In this paper, we describe an evolutionary attack and show that the adversarial examples produced by the attack have different features than those from gradient-based attacks. We also show that these features mean that one of the state-of-the-art defenses fails to detect such attacks.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2021

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

2021 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN)

ISBN

978-0-7381-3366-9

ISSN

2161-4393

e-ISSN

—

Počet stran výsledku

8

Strana od-do

—

Název nakladatele

IEEE

Místo vydání

NEW YORK

Místo konání akce

Online

Datum konání akce

18. 7. 2021

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000722581703104