AdvEdge: Optimizing Adversarial Perturbations Against Interpretable Deep Learning

  • Eldor Abdukhamidov
  • , Mohammed Abuhamad
  • , Firuz Juraev
  • , Eric Chan-Tin
  • , Tamer AbuHmed

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

11 Scopus citations

Abstract

Deep Neural Networks (DNNs) have achieved state-of-the-art performance in various applications. It is crucial to verify that the high accuracy prediction for a given task is derived from the correct problem representation and not from the misuse of artifacts in the data. Hence, interpretation models have become a key ingredient in developing deep learning models. Utilizing interpretation models enables a better understanding of how DNN models work, and offers a sense of security. However, interpretations are also vulnerable to malicious manipulation. We present AdvEdge and AdvEdge +, two attacks to mislead the target DNNs and deceive their combined interpretation models. We evaluate the proposed attacks against two DNN model architectures coupled with four representatives of different categories of interpretation models. The experimental results demonstrate our attacks’ effectiveness in deceiving the DNN models and their interpreters.

Original languageEnglish
Title of host publicationComputational Data and Social Networks - 10th International Conference, CSoNet 2021, Proceedings
EditorsDavid Mohaisen, Ruoming Jin
PublisherSpringer Science and Business Media Deutschland GmbH
Pages93-105
Number of pages13
ISBN (Print)9783030914332
DOIs
StatePublished - 2021
Event10th International Conference on Computational Data and Social Networks, CSoNet 2021 - Virtual Online
Duration: 15 Nov 202117 Nov 2021

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume13116 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference10th International Conference on Computational Data and Social Networks, CSoNet 2021
CityVirtual Online
Period15/11/2117/11/21

Keywords

  • Adversarial image
  • Deep learning
  • Interpretability

Fingerprint

Dive into the research topics of 'AdvEdge: Optimizing Adversarial Perturbations Against Interpretable Deep Learning'. Together they form a unique fingerprint.

Cite this