Back to articles
Articles
Volume: 32 | Article ID: art00013
Image
The Effect of Class Definitions on the Transferability of Adversarial Attacks Against Forensic CNNs
  DOI :  10.2352/ISSN.2470-1173.2020.4.MWSF-119  Published OnlineJanuary 2020
Abstract

In recent years, convolutional neural networks (CNNs) have been widely used by researchers to perform forensic tasks such as image tampering detection. At the same time, adversarial attacks have been developed that are capable of fooling CNN-based classifiers. Understanding the transferability of adversarial attacks, i.e. an attacks ability to attack a different CNN than the one it was trained against, has important implications for designing CNNs that are resistant to attacks. While attacks on object recognition CNNs are believed to be transferrable, recent work by Barni et al. has shown that attacks on forensic CNNs have difficulty transferring to other CNN architectures or CNNs trained using different datasets. In this paper, we demonstrate that adversarial attacks on forensic CNNs are even less transferrable than previously thought even between virtually identical CNN architectures! We show that several common adversarial attacks against CNNs trained to identify image manipulation fail to transfer to CNNs whose only difference is in the class definitions (i.e. the same CNN architectures trained using the same data). We note that all formulations of class definitions contain the unaltered class. This has important implications for the future design of forensic CNNs that are robust to adversarial and anti-forensic attacks.

Subject Areas :
Views 12
Downloads 0
 articleview.views 12
 articleview.downloads 0
  Cite this article 

Xinwei Zhao, Matthew C. Stamm, "The Effect of Class Definitions on the Transferability of Adversarial Attacks Against Forensic CNNsin Proc. IS&T Int’l. Symp. on Electronic Imaging: Media Watermarking, Security, and Forensics,  2020,  pp 119-1 - 119-7,  https://doi.org/10.2352/ISSN.2470-1173.2020.4.MWSF-119

 Copy citation
  Copyright statement 
Copyright © Society for Imaging Science and Technology 2020
72010604
Electronic Imaging
2470-1173
Society for Imaging Science and Technology
7003 Kilworth Lane, Springfield, VA 22151 USA