Exploiting Adversarial Embeddings for Better Steganography

This work proposes a protocol to iteratively build a distortion function for adaptive steganography while increasing its practical security after each iteration. It relies on prior art on targeted attacks and iterative design of steganalysis schemes. It combines targeted attacks on a given detector with a \min\max strategy, which dynamically selects the most difficult stego content associated with the best classifier at each iteration. We theoretically prove the convergence, which is confirmed by the practical results. Applied on J-Uniward this new protocol increases \perr from 7% to 20% estimated by Xu-Net, and from 10% to 23% for a non-targeted steganalysis by a linear classifier with GFR features.

[1]  Bin Li,et al.  Automatic Steganographic Distortion Learning Using a Generative Adversarial Network , 2017, IEEE Signal Processing Letters.

[2]  Marc Chaumont,et al.  Deep learning is a good steganalysis tool when embedding key is reused for different images, even if there is a cover sourcemismatch , 2015, Media Watermarking, Security, and Forensics.

[3]  Jessica J. Fridrich,et al.  Design of adaptive steganographic schemes for digital images , 2011, Electronic Imaging.

[4]  Jessica J. Fridrich,et al.  Universal distortion function for steganography in an arbitrary domain , 2014, EURASIP Journal on Information Security.

[5]  Yuan Yu,et al.  TensorFlow: A system for large-scale machine learning , 2016, OSDI.

[6]  Tomás Pevný,et al.  Exploring Non-Additive Distortion in Steganography , 2018, IH&MMSec.

[7]  Patrick Bas,et al.  Steganography via cover-source switching , 2016, 2016 IEEE International Workshop on Information Forensics and Security (WIFS).

[8]  Yi Zhang,et al.  Steganalysis of Adaptive JPEG Steganography Using 2D Gabor Filters , 2015, IH&MMSec.

[9]  Mo Chen,et al.  Deep Residual Network for Steganalysis of Digital Images , 2019, IEEE Transactions on Information Forensics and Security.

[10]  Guanshuo Xu,et al.  Deep Convolutional Neural Network to Detect J-UNIWARD , 2017, IH&MMSec.

[11]  Yun Q. Shi,et al.  Structural Design of Convolutional Neural Networks for Steganalysis , 2016, IEEE Signal Processing Letters.

[12]  Jiangqun Ni,et al.  Deep Learning Hierarchical Representations for Image Steganalysis , 2017, IEEE Transactions on Information Forensics and Security.

[13]  Jessica J. Fridrich,et al.  Low-Complexity Features for JPEG Steganalysis Using Undecimated DCT , 2015, IEEE Transactions on Information Forensics and Security.

[14]  Tomás Pevný,et al.  Using High-Dimensional Image Models to Perform Highly Undetectable Steganography , 2010, Information Hiding.

[15]  Jessica J. Fridrich,et al.  Gibbs Construction in Steganography , 2010, IEEE Transactions on Information Forensics and Security.

[16]  Tomás Pevný,et al.  "Break Our Steganographic System": The Ins and Outs of Organizing BOSS , 2011, Information Hiding.

[17]  Bin Li,et al.  CNN-Based Adversarial Embedding for Image Steganography , 2019, IEEE Transactions on Information Forensics and Security.

[18]  Jessica J. Fridrich,et al.  Minimizing embedding impact in steganography using trellis-coded quantization , 2010, Electronic Imaging.

[19]  Phil Sallee,et al.  Model-Based Methods For Steganography And Steganalysis , 2005, Int. J. Image Graph..

[20]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[21]  Jessica J. Fridrich,et al.  Content-Adaptive Steganography by Minimizing Statistical Detectability , 2016, IEEE Transactions on Information Forensics and Security.

[22]  Tomás Pevný,et al.  Is ensemble classifier needed for steganalysis in high-dimensional feature spaces? , 2015, 2015 IEEE International Workshop on Information Forensics and Security (WIFS).

[23]  Marc Chaumont,et al.  Adaptive steganography by oracle (ASO) , 2013, 2013 IEEE International Conference on Multimedia and Expo (ICME).

[24]  Jonathon Shlens,et al.  Explaining and Harnessing Adversarial Examples , 2014, ICLR.