UNCOVER: Development of an efficient steganalysis framework for uncovering hidden data in digital media

Vaila Leask (1) , Rémi Cogranne (2) , Dirk Borghys (1) , Helena Bruyninckx (1)

1 Royal Military Academy (RMA)
2 LM2S - Laboratoire Modélisation et Sûreté des Systèmes

This paper presents the general goals of Horizon 2020 project UNCOVER , whose overall purpose is to close the gap between academic work and operational needs in the fields of data-hiding. While digital data-hiding is a relatively new area of research, our motivation in this project has been rooted in the growing gap between the academic community and the operational needs of a "real-life" scenario of object inspection in order to UNCOVER the presence of data secretly hidden. As well as an oversight into the structure of UNCOVER, our paper presents an empirical study on the impact of specifically training a detection method for a given data-hiding scheme, the socalled Stego-Source Mismatch, as an example of unexplored issues that raises important and mostly ignored consequences within the operational context the UNCOVER project targets.

Mots clés: Image processing; Security; Data-hiding; Contest; H2020; Steganography; Steganalysis; Forensics; Cover-Source Mismatch

Domaines: Cryptographie et sécurité [cs.CR]; Traitement du signal et de l'image [eess.SP]

Vaila Leask, Rémi Cogranne, Dirk Borghys, Helena Bruyninckx. UNCOVER: Development of an efficient steganalysis framework for uncovering hidden data in digital media. 17th International Conference on Availability, Reliability and Security (ARES 2022), Aug 2022, Vienna, Austria. ⟨hal-03696116⟩