Making targeted black-box evasion attacks effective and efficient

Tutkimustuotos: Artikkeli kirjassa/konferenssijulkaisussaKonferenssiesitysvertaisarvioitu

Tutkijat

Organisaatiot

Kuvaus

We investigate how an adversary can optimally use its query budget for targeted evasion attacks against deep neural networks in a black-box setting. We formalize the problem setting and systematically evaluate what benefits the adversary can gain by using substitute models. We show that there is an exploration-exploitation tradeoff in that query efficiency comes at the cost of effectiveness. We present two new attack strategies for using substitute models and show that they are as effective as previous query-only techniques but require significantly fewer queries, by up to three orders of magnitude. We also show that an agile adversary capable of switching through different attack techniques can achieve pareto-optimal efficiency. We demonstrate our attack against Google Cloud Vision showing that the difficulty of black-box attacks against real-world prediction APIs is significantly easier than previously thought (requiring approximately 500 queries instead of approximately 20,000 as in previous works).

Yksityiskohdat

AlkuperäiskieliEnglanti
Sivumäärä12
TilaHyväksytty/In press - 12 elokuuta 2019
OKM-julkaisutyyppiEi oikeutettu
TapahtumaACM Workshop on Artificial Intelligence and Security - London, Iso-Britannia
Kesto: 15 marraskuuta 201915 marraskuuta 2019
Konferenssinumero: 12
https://aisec.cc/

Workshop

WorkshopACM Workshop on Artificial Intelligence and Security
LyhennettäAISec
MaaIso-Britannia
KaupunkiLondon
Ajanjakso15/11/201915/11/2019
www-osoite

ID: 36220015