Facing the cover-source mismatch on JPHide using training-set design

Dirk Borghys, Helena Bruyninckx, Patrick Bas

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

Abstract

This short paper investigates the influence of the image processing pipeline (IPP) on the cover-source mismatch (CSM) for the popular JPHide steganographic scheme. We propose to deal with CSM by combining a forensics and a steganalysis approach. A multiclassifier is first trained to identify the IPP, and secondly a specific training set is designed to train a targeted classifier for steganalysis purposes. We show that the forensic step is immune to the steganographic embedding. The proposed IPP-informed steganalysis outperforms classical strategies based on training on a mixture of sources and we show that it can provide results close to a detector specifically trained on the appropriate source.

Original languageEnglish
Title of host publicationIH and MMSec 2018 - Proceedings of the 6th ACM Workshop on Information Hiding and Multimedia Security
PublisherAssociation for Computing Machinery, Inc
Pages17-22
Number of pages6
ISBN (Electronic)9781450356251
DOIs
Publication statusPublished - 14 Jun 2018
Event6th ACM Workshop on Information Hiding and Multimedia Security, IH and MMSec 2018 - Innsbruck, Austria
Duration: 20 Jun 201822 Jun 2018

Publication series

NameIH and MMSec 2018 - Proceedings of the 6th ACM Workshop on Information Hiding and Multimedia Security

Conference

Conference6th ACM Workshop on Information Hiding and Multimedia Security, IH and MMSec 2018
Country/TerritoryAustria
CityInnsbruck
Period20/06/1822/06/18

Fingerprint

Dive into the research topics of 'Facing the cover-source mismatch on JPHide using training-set design'. Together they form a unique fingerprint.

Cite this