Extracting relevant features of steganographic schemes by feature selection techniques - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2007

Extracting relevant features of steganographic schemes by feature selection techniques

Résumé

This paper analyses the security of dirty paper trellis (DPT) watermarking schemes which use both informed coding and informed em- bedding. After recalling the principles of message embedding with DPT watermarking, the secret parameters of the scheme are highlighted. The security weaknesses of DPT watermarking are then presented: in the wa- termarked contents only attack (WOA) setup, the watermarked data-set exhibits clusters corresponding to the different patterns attached to the arcs of the trellis. The K-means clustering algorithm is used to estimate these patterns and a co-occurrence analysis is performed to retrieve the connectivity of the trellis. Experimental results demonstrate that it is possible to accurately estimate the trellis configuration, which enables to perform attacks much more efficient than simple additive white Gaus- sian noise (AWGN).
Fichier principal
Vignette du fichier
paperWachaMiche.pdf (232.47 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-00166700 , version 1 (08-08-2007)

Identifiants

  • HAL Id : hal-00166700 , version 1

Citer

Yoan Miche, Patrick Bas, Christian Jutten, Amaury Lendasse, Olli Simula. Extracting relevant features of steganographic schemes by feature selection techniques. WaCHa 2007 - Third Wavila Challenge, Jun 2007, Saint-Malo, France. pp.electronic version. ⟨hal-00166700⟩
295 Consultations
163 Téléchargements

Partager

Gmail Facebook X LinkedIn More