The Cover Source Mismatch Problem in Deep-Learning Steganalysis

Giboulot Quentin, Bas Patrick, Cogranne Rémi, Borghys Dirk

Onderzoeksoutput: Hoofdstuk in Boek/Rapport/CongresprocedureConferentiebijdragepeer review

Samenvatting

This paper studies the problem of Cover Source Mismatch (CSM) in steganalysis, i.e. the impact of a testing set which does not originate from the same source than the training set. In this study, the trained steganalyzer uses state of the art deep-learning architecture prone to better generalization than feature-based steganalysis. Different sources such as the sensor model, the ISO sensitivity, the processing pipeline and the content, are investigated. Our conclusions are that, on one hand, deep learning steganalysis is still very sensitive to the CSM, on the other hand, the holistic strategy leverages the good generalization properties of deep learning to reduce the CSM with a relatively small number of training samples.

Originele taal-2Engels
Titel30th European Signal Processing Conference, EUSIPCO 2022 - Proceedings
UitgeverijEuropean Signal Processing Conference, EUSIPCO
Pagina's1032-1036
Aantal pagina's5
ISBN van elektronische versie9789082797091
StatusGepubliceerd - 2022
Evenement30th European Signal Processing Conference, EUSIPCO 2022 - Belgrade, Servië
Duur: 29 aug. 20222 sep. 2022

Publicatie series

NaamEuropean Signal Processing Conference
Volume2022-August
ISSN van geprinte versie2219-5491

Congres

Congres30th European Signal Processing Conference, EUSIPCO 2022
Land/RegioServië
StadBelgrade
Periode29/08/222/09/22

Vingerafdruk

Duik in de onderzoeksthema's van 'The Cover Source Mismatch Problem in Deep-Learning Steganalysis'. Samen vormen ze een unieke vingerafdruk.

Citeer dit