Image Steganography Scheme Based on GANs

doi:10.3969/j.issn.1671-1122.2019.05.007

Abstract

Abstract:

Focused on the issue that information-hidden carriers will leave traces of modification, and it is fundamentally difficult to resist the statistical steganalysis algorithm detection. A new security steganography model based on Generative Adversarial Networks (GAN) is proposed.By generating more appropriate carrier information to ensure the security of information hiding and improving the imperceptibility, we can effectively resist the steganalysis algorithm detection.The proposed strategy is to jointly generate the anti-network and traditional information hiding, and use the “generated” image and the “traditional” image as the discriminator input, and iteratively optimizes it,which make the image generated by the generator more suitable as carrier information. Compared with SGAN and SSGAN, this model reduces the detection accuracy of steganalysis by 9% and 4.1% respectively. Experimental results show that the new scheme is superior to the contrast algorithm in Anti-steganography analysis and security indicators.

Key words: steganography, GAN, WGAN, security

CLC Number:

TP309

Yaojie WANG, Ke NIU, Xiaoyuan YANG. Image Steganography Scheme Based on GANs[J]. Netinfo Security, 2019, 19(5): 54-60.

Figures/Tables 8

References 20

[1]	XIONG Jinbo, MA Rong, ZHANG Yuanyuan, et al.Image Information Hiding Method and Implementation for Social Network[J]. Information Network Security, 2017, 17(3): 6-13.
	熊金波,马蓉,张媛媛,等.面向社交网络的图片信息隐藏方法与实现[J].信息网络安全,2017,17(3):6-13.
[2]	TIEKEL A Z, RANKIN G A, SCHYNDEL R V.Electronic Watermark[C]// IEEE. Digital Image Computing, Technology and Applications, December 10-14, 1993, Sydney, Australia. New York: IEEE, 1993: 666-673.
[3]	YANG C H, WENG C Y, WANG S J.Adaptive Data Hiding in Edge Areas of Images with Spatial LSB Domain Systems[J]. IEEE Transactions on Information Forensics & Security, 2008, 3(3): 488-497.
[4]	GAO Hongtao, LU Wei, YANG Yuwang.Image Saliency Region Extraction Method Based on Adaptive Adjustment Kernel Function[J]. Netinfo Security, 2018, 18(2): 54-60.
	高洪涛,陆伟,杨余旺.基于自适应调节核函数的图像显著区域提取方法[J].信息网络安全,2018,18(2):54-60.
[5]	PEVNY T, FILLER T, BAS P. Using High-Dimensional Image Models to Perform Highly Undetectable Steganography[EB/OL]. , 2017-8-15.
[6]	HOLUB V, FRIDRICH J.Designing Steganographic Distortion Using Directional Filters[C]//IEEE. International Workshop on Information Forensics and Security, December1-5, 2013, Chicago, USA. New York: IEEE, 2013: 234-239.
[7]	RUANAIDH J K O, Dowling W J, Boland F M. Phase Watermarking of Digital Images[C]//IEEE. 2013 International Conference on Image Processing, May 11-14, 2013, San Francisco, USA. New York: IEEE, 2013: 1673-1687.
[8]	COX I J, KILIAN J, LEIGHTON F T.Secure Spread Spectrum Watermarking for Multimedia[J]. IEEE Transactions on Image Processing, 2010, 6(12): 1673-1687.
[9]	LIN W H, HORNY S J, KAO T W.An Efficient Watermarking Method Based on Significant Difference of Wavelet Coefficient Quantization[J]. IEEE Transactions on Multimedia, 2008, 10(5): 746-757.
[10]	DENIS V, BORIS B, BUENAIV E.Generative Adversarial Networks for Image Steganography[R]. ICLR 2016 Open Review, 2016.
[11]	SHI H, DONG J, WANG W, et al. SSGAN: Secure Steganography Based on Generative Adversarial Networks[EB/OL]. , 2017-7-30.
[12]	GOODFELLOW I, POUGAT A J, MIRZA M. Generative Adversarial Networks[EB/OL]. , 2014-6-10.
[13]	WANG Kunfeng, YU Chao, DUAN Yanjie, et al.Research Progress and Prospects of Generative Confrontation Network GAN[J]. Acta Automatica Sinica, 2017, 43(3): 321-332.
	王坤峰,苟超,段艳杰,等.生成式对抗网络GAN的研究进展与展望[J].自动化学报,2017,43(3):321-332.
[14]	RADFORD A, METZ L, CHINTALA S.Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks[J]. Computer Science, 2015, 1(11): 132-141.
[15]	MIRZA M, OSINDRO S. Conditional Generative Adversarial Nets[EB/OL]. , 2014-11-30.
[16]	CHEN X, DUAN Y, HOUTHOOFT R, et al. InfoGAN: Interpretable Representation Learning by Information MaximiZing Generative Adversarial Nets[EB/OL]. , 2016-6-30.
[17]	ARJOVSKY M, BOTTOU L. Towards Principled Methods for Training Generative Adversarial Networks[EB/OL]. . 04862, 2017-2-9.
[18]	CUI S, JIANG Y.Effective Lipschitz Constraint Enforcement for Wasserstein GAN Training[C]//IEEE. 2017 2nd IEEE International Conference on Computational Intelligence and Applications (ICCIA) IEEE, September 8-11, 2017, Beijing, China. New York: IEEE, 2017: 74-78.
[19]	QIAN Y, DONG J, WANG W, et al. Deep Learning for Steganalysis via Convolutional Neural Networks[EB/OL]. , 2015-4-30.
[20]	ACHAB M, GUILLOUX A, S G, et al. SGD with Variance Reduction beyond Empirical Risk Minimization[EB/OL]. , 2015-10-30.

样本类型	SGAN	SSGAN	Improved-SSGAN
传统式载密图像/%	0.117
生成式载密图像/%	0.183	0.232	0.273