Skip to main navigation Skip to search Skip to main content

Detection of Adversarial Attacks and Characterization of Adversarial Subspace

  • École de technologie supérieure

Research output: Contribution to Book/Report types › Contribution to conference proceedings › peer-review

20 Citations (Scopus)

Abstract

Adversarial attacks have always been a serious threat for any data-driven model. In this paper, we explore subspaces of adversarial examples in unitary vector domain, and we propose a novel detector for defending our models trained for environmental sound classification. We measure chordal distance between legitimate and malicious representation of sounds in unitary space of generalized Schur decomposition and show that their manifolds lie far from each other. Our front-end detector is a regularized logistic regression which discriminates eigenvalues of legitimate and adversarial spectrograms. The experimental results on three benchmarking datasets of environmental sounds represented by spectrograms reveal high detection rate of the proposed detector for eight types of adversarial attacks and it also outperforms other detection approaches.

Original languageEnglish
Title of host publication2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages3097-3101
Number of pages5
ISBN (Electronic)9781509066315
DOIs
Publication statusPublished - May 2020
Event2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Barcelona, Spain
Duration: 4 May 2020 → 8 May 2020

Publication series

NameICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Volume2020-May
ISSN (Print)1520-6149

Conference

Conference2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020
Country/TerritorySpain
CityBarcelona
Period4/05/20 → 8/05/20

Keywords

  • Generalized Schur decomposition
  • adversarial detection.
  • adversarial subspace
  • chordal distance

ASJC Scopus subject areas

  • Software
  • Signal Processing
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Detection of Adversarial Attacks and Characterization of Adversarial Subspace'. These topics are generated from the title and abstract of the publication. Together, they form a unique fingerprint.

Cite this