Universal distortion function for steganography in an arbitrary domain

By Vojtech Holub, Jessica J. Fridrich, Tomáš Denemark. EURASIP Journal on Information Security, 2014

1,115 citations recorded by the source corpus.

Currently, the most successful approach to steganography in empirical objects, such as digital media, is to embed the payload while minimizing a suitably defined distortion function. The design of the distortion is essentially the only task left to the steganographer since efficient practical codes exist that embed near the payload-distortion bound. The practitioner’s goal is to design the distortion to obtain a scheme with a high empirical statistical detectability. In this paper, we propose a universal distortion design called universal wavelet relative distortion (UNIWARD) that can be applied for embedding in an arbitrary domain. The embedding distortion is computed as a sum of relative changes of coefficients in a directional filter bank decomposition of the cover image. The directionality forces the embedding changes to such parts of the cover object that are difficult to model in multiple directions, such as textures or noisy regions, while avoiding smooth regions or clean edges. We demonstrate experimentally using rich models as well as targeted attacks that steganographic methods built using UNIWARD match or outperform the current state of the art in the spatial domain, JPEG domain, and side-informed JPEG domain.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/e2a8598046b35101f134ed7350196ce937041421

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Yedroudj-Net: An Efficient CNN for Spatial Steganalysis (2018, positive): “In Table 1, we report the error probability obtained when stegana-lyzing WOW and S-UNIWARD embedding algorithms at 0.2 bpp and 0.4 bpp.”
Automatic Steganographic Distortion Learning Using a Generative Adversarial Network (2017, methodology): “[29] T. Denemark, J. Fridrich, and V. Holub, “Further study on the security of S-UNIWARD,” in Proc.”
SteganoGAN: High Capacity Image Steganography with GANs (2019, methodology): “Finally, we compare the detection error for images generated by STEGANOGAN against those reported by (Ye et al., 2017) on images generated by three state-of-the-art steganography algorithms: WOW (Holub & Fridrich, 2012), S-UNIWARD (Holub et al., 2014), and HILL (Li et al., 2014).”
Content-Adaptive Steganography by Minimizing Statistical Detectability (2016, methodology): “We selected four content-adaptive steganographic techniques that appear to be the state of the art as of writing this paper (April 2015): WOW [3], S-UNIWARD implemented with the stabilizing constant σ = 1 as described in [4], HUGO-BD [2] implemented using the Gibbs construction with”
Depth-Wise Separable Convolutions and Multi-Level Pooling for an Efficient Spatial CNN-Based Steganalysis (2020, methodology): “For a fair comparison, we report the performance of Yedroudj-Net, Ye-Net, Xu-Net, SRNet and the Spatial Rich Model + the Ensemble Classifier (SRM + EC), against the embedding algorithm WOW and S-UNIWARD at payload 0.2 bpp and 0.4 bpp.”
E2Style: Improve the Efficiency and Effectiveness of StyleGAN Inversion (2021, background): “In contrast to traditional steganography [67], [68], [69], which embeds secret messages with small capacity, deep hiding aims at large capacity hidden messages, such as hiding a secret image with 1024 × 1024 resolution into a carrier image.”
Deep Residual Network for Steganalysis of Digital Images (2019, methodology): “5 bpp (bits per pixel) for WOW [28], HILL [41], and S-UNIWARD [31].”
Generating steganographic images via adversarial training (2017, background): “Finally, S-UNIWARD [10] proposes a universal distortion function that is agnostic to the embedding domain.”
Deep residual learning for image steganalysis (2018, methodology): “Four states of the art steganographic algorithms, including the Wavelet ObtainedWeights steganography (WOW) [14], S-UNIWARD [13], HILL [19] and MiPOD [29], are used for evaluating the effectiveness of steganalytic algorithms.”
HiDDeN: Hiding Data With Deep Networks (2018, methodology): “S-UNIWARD [11] is similar to WOW but can be used for embedding in an arbitrary domain.”