Transferable perturbations of deep feature distributions

N Inkawhich, KJ Liang, L Carin

Research output: Contribution to journalArticlepeer-review

43 Scopus citations

Abstract

Almost all current adversarial attacks of CNN classifiers rely on information derived from the output layer of the network. This work presents a new adversarial attack based on the modeling and exploitation of class-wise and layer-wise deep feature distributions. We …
Original languageEnglish (US)
JournalarXiv preprint arXiv:2004.12519
StatePublished - 2020
Externally publishedYes

Fingerprint

Dive into the research topics of 'Transferable perturbations of deep feature distributions'. Together they form a unique fingerprint.

Cite this