ABSTRACT
In comparison to symmetric embedding, asymmetric methods generally provide better steganography security. However, the performance of existing asymmetric methods is limited by their reliance on symmetric embedding costs. In this paper, we present a novel Generative Adversarial Network (GAN)-based steganography approach that independently learns asymmetric embedding costs from scratch. Our proposed framework features a generator with a dual-branch architecture and a discriminator that integrates multiple steganalytic networks. To address the issues of model instability and non-convergence that often arise in GAN model training, we implement an adaptive strategy that updates the GAN model parameters according to the performance of multiple steganalytic networks in each iteration. Furthermore, we introduce a new adversarial loss function that effectively learns asymmetric embedding costs by utilizing features like image residuals, gradients, asymmetric embedding probability maps, and the sign of the modification map to train the dual-branch network within the generator. Our comprehensive experiments show that our method achieves state-of-the-art steganography security results, significantly outperforming existing top-performing symmetric and asymmetric methods. Additionally, numerous ablation experiments confirm the rationality of our GAN-based model design.
- Patrick Bas, Tomás Filler, and Tomás Pevný. 2011. Break our steganographic system: the ins and outs of organizing BOSS. International Workshop on Information Hiding (2011), 59--70.Google ScholarDigital Library
- Solè ne Bernard, Patrick Bas, John Klein, and Tomá s Pevný. 2020. Explicit optimization of min max steganographic game. IEEE Transactions on Information Forensics and Security, Vol. 16 (2020), 812--823.Google ScholarCross Ref
- Solè ne Bernard, Patrick Bas, Tomá s Pevný, and John Klein. 2021. Optimizing Additive Approximations of Non-additive Distortion Functions. In ACM Workshop on Information Hiding and Multimedia Security. ACM, 105--112.Google Scholar
- Mehdi Boroumand, Mo Chen, and Jessica J. Fridrich. 2018. Deep residual network for steganalysis of digital images. IEEE Transactions on Information Forensics and Security, Vol. 14, 5 (2018), 1181--1193.Google ScholarDigital Library
- Yunjey Choi, Min-Je Choi, Munyoung Kim, Jung-Woo Ha, Sunghun Kim, and Jaegul Choo. 2018. StarGAN: Unified generative adversarial networks for multi-domain image-to-image translation. In IEEE Conference on Computer Vision and Pattern Recognition. IEEE, 8789--8797.Google ScholarCross Ref
- Rémi Cogranne, Quentin Giboulot, and Patrick Bas. 2022. Efficient Steganography in JPEG Images by Minimizing Performance of Optimal Detector. IEEE Transactions on Information Forensics and Security, Vol. 17 (2022), 1328--1343.Google ScholarCross Ref
- Xiaoqing Deng, Bolin Chen, Weiqi Luo, and Da Luo. 2019. Fast and effective global covariance pooling network for image steganalysis. In ACM Workshop on Information Hiding and Multimedia Security. ACM, 230--234.Google ScholarDigital Library
- Tomás Filler, Jan Judas, and Jessica J. Fridrich. 2011. Minimizing additive distortion in steganography using syndrome-trellis codes. IEEE Transactions on Information Forensics and Security, Vol. 6, 3 (2011), 920--935.Google ScholarDigital Library
- Jessica J. Fridrich and Tomás Filler. 2007. Practical methods for minimizing embedding impact in steganography. In Security, Steganography, and Watermarking of Multimedia Contents. SPIE, 650502.Google Scholar
- Jessica J. Fridrich and Jan Kodovský. 2012. Rich models for steganalysis of digital images. IEEE Transactions on Information Forensics and Security, Vol. 7, 3 (2012), 868--882.Google ScholarDigital Library
- Ian J. Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron C. Courville, and Yoshua Bengio. 2014. Generative adversarial nets. In Neural Information Processing Systems. ACM, 2672--2680.Google Scholar
- Vojtech Holub, Jessica J. Fridrich, and Tomás Denemark. 2014. Universal distortion function for steganography in an arbitrary domain. EURASIP Journal on Information Security, Vol. 2014 (2014), 1--13.Google ScholarCross Ref
- Ying Jin, Yunbo Wang, Mingsheng Long, Jianmin Wang, Philip S. Yu, and Jiaguang Sun. 2020. A multi-player minimax game for generative adversarial networks. In IEEE International Conference on Multimedia and Expo. IEEE, 1--6.Google ScholarCross Ref
- Bin Li, Shunquan Tan, Ming Wang, and Jiwu Huang. 2014a. Investigation on Cost Assignment in Spatial Image Steganography. IEEE Transactions on Information Forensics and Security, Vol. 9, 8 (2014), 1264--1277.Google ScholarDigital Library
- Bin Li, Ming Wang, Jiwu Huang, and Xiaolong Li. 2014b. A new cost function for spatial image steganography. In IEEE International Conference on Image Processing. IEEE, 4206--4210.Google ScholarCross Ref
- Bin Li, Ming Wang, Xiaolong Li, Shunquan Tan, and Jiwu Huang. 2015. A strategy of clustering modification directions in spatial image steganography. IEEE Transactions on Information Forensics and Security, Vol. 10, 9 (2015), 1905--1917.Google ScholarDigital Library
- Weixiang Li, Weiming Zhang, Li Li, Hang Zhou, and Nenghai Yu. 2020. Designing near-optimal steganographic codes in practice based on polar codes. IEEE Transactions on Communications, Vol. 68, 7 (2020), 3948--3962.Google ScholarCross Ref
- Minglin Liu, Weiqi Luo, Peijia Zheng, and Jiwu Huang. 2021. A new adversarial embedding method for enhancing image steganography. IEEE Transactions on Information Forensics and Security, Vol. 16 (2021), 4621--4634.Google ScholarCross Ref
- Minglin Liu, Tingting Song, Weiqi Luo, Peijia Zheng, and Jiwu Huang. 2023. Adversarial steganography embedding via stego generation and selection. IEEE Transactions on Dependable and Secure Computing, Vol. 20, 3 (2023), 2375--2389.Google ScholarDigital Library
- Huaxiao Mo, Tingting Song, Bolin Chen, Weiqi Luo, and Jiwu Huang. 2019. Enhancing JPEG steganography using iterative adversarial examples. In IEEE International Workshop on Information Forensics and Security. IEEE, 1--6.Google ScholarCross Ref
- Xianbo Mo, Shunquan Tan, Weixuan Tang, Bin Li, and Jiwu Huang. 2023. ReLOAD: Using Reinforcement Learning to Optimize Asymmetric Distortion for Additive Steganography. IEEE Transactions on Information Forensics and Security, Vol. 18 (2023), 1524--1538.Google ScholarDigital Library
- Xinghong Qin, Bin Li, Shunquan Tan, Weixuan Tang, and Jiwu Huang. 2022. Gradually Enhanced Adversarial Perturbations on Color Pixel Vectors for Image Steganography. IEEE Transactions on Circuits and Systems for Video Technology, Vol. 32, 8 (2022), 5110--5123.Google ScholarDigital Library
- Olaf Ronneberger, Philipp Fischer, and Thomas Brox. 2015. U-net: Convolutional networks for biomedical image segmentation. In Medical Image Computing and Computer-Assisted Intervention, Vol. 9351. Springer, 234--241.Google Scholar
- Vahid Sedighi, Ré mi Cogranne, and Jessica J. Fridrich. 2015. Content-adaptive steganography by minimizing statistical detectability. IEEE Transactions on Information Forensics and Security, Vol. 11, 2 (2015), 221--234.Google ScholarDigital Library
- Tingting Song, Minglin Liu, Weiqi Luo, and Peijia Zheng. 2021. Enhancing image steganography via stego generation and selection. In IEEE International Conference on Acoustics, Speech and Signal Processing. IEEE, 2695--2699.Google ScholarCross Ref
- Weixuan Tang, Bin Li, Mauro Barni, Jin Li, and Jiwu Huang. 2021. An automatic cost learning framework for image steganography using deep reinforcement learning. IEEE Transactions on Information Forensics and Security, Vol. 16 (2021), 952--967.Google ScholarCross Ref
- Weixuan Tang, Bin Li, Shunquan Tan, Mauro Barni, and Jiwu Huang. 2019. CNN-based adversarial embedding for image steganography. IEEE Transactions on Information Forensics and Security, Vol. 14, 8 (2019), 2074--2087.Google ScholarDigital Library
- Weixuan Tang, Shunquan Tan, Bin Li, and Jiwu Huang. 2017. Automatic steganographic distortion learning using a generative adversarial network. IEEE Signal Processing Letters, Vol. 24, 10 (2017), 1547--1551.Google ScholarCross Ref
- Yi-Lun Wu, Hong-Han Shuai, Zhi Rui Tam, and Hong-Yu Chiu. 2021. Gradient normalization for generative adversarial networks. In IEEE International Conference on Computer Vision. IEEE, 6353--6362.Google ScholarCross Ref
- Guanshuo Xu, Han-Zhou Wu, and Yun-Qing Shi. 2016. Structural design of convolutional neural networks for steganalysis. IEEE Signal Processing Letters, Vol. 23, 5 (2016), 708--712.Google ScholarCross Ref
- Jianhua Yang, Danyang Ruan, Jiwu Huang, Xiangui Kang, and Yun-Qing Shi. 2020. An embedding cost learning framework using GAN. IEEE Transactions on Information Forensics and Security, Vol. 15 (2020), 839--851.Google ScholarDigital Library
- Mehdi Yedroudj, Frédéric Comby, and Marc Chaumont. 2018. Yedroudj-Net: An efficient CNN for spatial steganalysis. In IEEE International Conference on Acoustics, Speech and Signal Processing. IEEE, 2092--2096.Google ScholarDigital Library
- Ru Zhang, Feng Zhu, Jianyi Liu, and Gongshen Liu. 2020. Depth-wise separable convolutions and multi-level pooling for an efficient spatial CNN-based steganalysis. IEEE Transactions on Information Forensics and Security, Vol. 15 (2020), 1138--1150.Google ScholarCross Ref
- Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A. Efros. 2017. Unpaired image-to-image translation using cycle-consistent adversarial networks. In IEEE International Conference on Computer Vision. IEEE, 2242--2251.Google Scholar
Index Terms
- Automatic Asymmetric Embedding Cost Learning via Generative Adversarial Networks
Recommendations
Enhancing Adversarial Embedding based Image Steganography via Clustering Modification Directions
Image steganography is a technique used to conceal secret information within cover images without being detected. However, the advent of convolutional neural networks (CNNs) has threatened the security of image steganography. Due to the inherent ...
Invisible steganography via generative adversarial networks
Nowadays, there are plenty of works introducing convolutional neural networks (CNNs) to the steganalysis and exceeding conventional steganalysis algorithms. These works have shown the improving potential of deep learning in information hiding domain. ...
GAN-based image steganography for enhancing security via adversarial attack and pixel-wise deep fusion
AbstractIn recent years, the development of steganalysis based on convolutional neural networks (CNN) has brought new challenges to the security of image steganography. However, the current steganographic methods are difficult to resist the detection of ...
Comments