Content-adaptive pentary steganography using the multivariate generalized Gaussian cover model

Abstract

The vast majority of steganographic schemes for digital images stored in the raster format limit the amplitude of embedding changes to the smallest possible value. In this paper, we investigate the possibility to further improve the empirical security by allowing the embedding changes in highly textured areas to have a larger amplitude and thus embedding there a larger payload. Our approach is entirely model driven in the sense that the probabilities with which the cover pixels should be changed by a certain amount are derived from the cover model to minimize the power of an optimal statistical test. The embedding consists of two steps. First, the sender estimates the cover model parameters, the pixel variances, when modeling the pixels as a sequence of independent but not identically distributed generalized Gaussian random variables. Then, the embedding change probabilities for changing each pixel by 1 or 2, which can be transformed to costs for practical embedding using syndrome-trellis codes, are computed by solving a pair of non-linear algebraic equations. Using rich models and selection-channel-aware features, we compare the security of our scheme based on the generalized Gaussian model with pentary versions of two popular embedding algorithms: HILL and S-UNIWARD.

Improving Steganographic Security by Synchronizing the Selection Channel

“…Since the empirical detectability does not change with an increased number of sweeps, the proposed scheme utilizes merely a single Gibbs sweep. We prove the usefulness of the framework by applying it to the costs of the (ternary) scheme proposed in [17] with the multivariate Gaussian cover model (MVG) and HILL [15]. In both cases, the empirical security is markedly improved when testing with the spatial rich model [5] (SRM) as well as its selection-channel aware version [1] (maxSRMd2).…”

“…As this distortion is not only non-additive but most importantly nonlocal, it is upper-bounded by another function that can be written as a sum of locally supported potentials and one that can be implemented using the Gibbs construction [2]. HUGO-BD's empirical security, however, is subpar when compared to current state-of-the-art additive schemes, such as S-UNIWARD [11], HILL [15], and the approach based on minimizing the detectability of an optimal detector within a chosen cover model [17].…”

Review of distortion function design in spatial domain

“…MVGG (Multivariate Generalized Gaussian) [20] use the similar framework with MG, and has two improvement. Firstly, MVGG explore the possibility to further improve the empirical security by using a more general cover model.…”

Hybrid Data Hiding Scheme Using Right-Most Digit Replacement and Adaptive Least Significant Bit for Digital Images

“…It consists of a high dimensional image model to calculate the distortions corresponding to a modification of each pixel by˘1, but can only hide up to 1 bpp. Similarly, wavelet obtained weights (WOW) [15], spatial universal wavelet relative distortion (S-UNIWARD) [16], HIgh Low Low (HILL) [17], and minimizing the power of optimal detector (MiPOD) [18] are targeted as highly secure steganographic methods with limited payload.…”