Skip to main navigation Skip to search Skip to main content

The Quantitative Relationship between Adversarial Training and Robustness of CNN Model

  • Jie Wang
  • , Minyan Lu
  • , Jun Ai*
  • , Xueyuan Sun
  • *Corresponding author for this work
  • Beihang University

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

Abstract

With the increasing application of deep neural networks in security-critical systems, robustness becomes an important property for deep learning. However, deep neural networks are very vulnerable to perturbations, especially adversarial attacks. The adversarial training by adding adversarial examples to the training set has become a common method to improve the performance of deep neural networks. In this paper, based on the existing robustness metrics and indicators, we study the quantitative influence of adversarial training on the robustness of network-in-network and residual network, and the differences of indicators are compared. The experimental results show that adversarial training can affect the accuracy and robustness of CNN models, and the variation trends of testing accuracy and robustness are opposite.

Original languageEnglish
Title of host publicationProceedings - 2020 7th International Conference on Dependable Systems and Their Applications, DSA 2020
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages543-549
Number of pages7
ISBN (Electronic)9780738124223
DOIs
StatePublished - Nov 2020
Event7th International Conference on Dependable Systems and Their Applications, DSA 2020 - Virtual, Xi�an, China
Duration: 28 Nov 202029 Nov 2020

Publication series

NameProceedings - 2020 7th International Conference on Dependable Systems and Their Applications, DSA 2020

Conference

Conference7th International Conference on Dependable Systems and Their Applications, DSA 2020
Country/TerritoryChina
CityVirtual, Xi�an
Period28/11/2029/11/20

Keywords

  • adversarial examples
  • deep neural network
  • global adversarial robustness
  • local adversarial robustness
  • robustness

Fingerprint

Dive into the research topics of 'The Quantitative Relationship between Adversarial Training and Robustness of CNN Model'. Together they form a unique fingerprint.

Cite this