Progressive selection-channel networks for image steganalysis
Tian Wu
Key Laboratory of Aerospace Information Security and Trusted Computing, School of Cyber Science and Engineering, Ministry of Education, Wuhan University, Wuhan, China
Search for more papers by this authorCorresponding Author
Lina Wang
Key Laboratory of Aerospace Information Security and Trusted Computing, School of Cyber Science and Engineering, Ministry of Education, Wuhan University, Wuhan, China
Correspondence Lina Wang and Liming Zhai, Key Laboratory of Aerospace Information Security and Trusted Computing, School of Cyber Science and Engineering, Ministry of Education, Wuhan University, Wuhan 430072, China.
Email: [email protected] and
Search for more papers by this authorCorresponding Author
Liming Zhai
Key Laboratory of Aerospace Information Security and Trusted Computing, School of Cyber Science and Engineering, Ministry of Education, Wuhan University, Wuhan, China
Correspondence Lina Wang and Liming Zhai, Key Laboratory of Aerospace Information Security and Trusted Computing, School of Cyber Science and Engineering, Ministry of Education, Wuhan University, Wuhan 430072, China.
Email: [email protected] and
Search for more papers by this authorCanming Fang
Key Laboratory of Aerospace Information Security and Trusted Computing, School of Cyber Science and Engineering, Ministry of Education, Wuhan University, Wuhan, China
Search for more papers by this authorMingcheng Zhang
Key Laboratory of Aerospace Information Security and Trusted Computing, School of Cyber Science and Engineering, Ministry of Education, Wuhan University, Wuhan, China
Search for more papers by this authorTian Wu
Key Laboratory of Aerospace Information Security and Trusted Computing, School of Cyber Science and Engineering, Ministry of Education, Wuhan University, Wuhan, China
Search for more papers by this authorCorresponding Author
Lina Wang
Key Laboratory of Aerospace Information Security and Trusted Computing, School of Cyber Science and Engineering, Ministry of Education, Wuhan University, Wuhan, China
Correspondence Lina Wang and Liming Zhai, Key Laboratory of Aerospace Information Security and Trusted Computing, School of Cyber Science and Engineering, Ministry of Education, Wuhan University, Wuhan 430072, China.
Email: [email protected] and
Search for more papers by this authorCorresponding Author
Liming Zhai
Key Laboratory of Aerospace Information Security and Trusted Computing, School of Cyber Science and Engineering, Ministry of Education, Wuhan University, Wuhan, China
Correspondence Lina Wang and Liming Zhai, Key Laboratory of Aerospace Information Security and Trusted Computing, School of Cyber Science and Engineering, Ministry of Education, Wuhan University, Wuhan 430072, China.
Email: [email protected] and
Search for more papers by this authorCanming Fang
Key Laboratory of Aerospace Information Security and Trusted Computing, School of Cyber Science and Engineering, Ministry of Education, Wuhan University, Wuhan, China
Search for more papers by this authorMingcheng Zhang
Key Laboratory of Aerospace Information Security and Trusted Computing, School of Cyber Science and Engineering, Ministry of Education, Wuhan University, Wuhan, China
Search for more papers by this authorAbstract
Steganalysis is a detection technology against steganography that embeds secret data into digital media carriers. The selection channel, which indicates the embedding details of steganography, is well recognized in boosting the detection performance of image steganalysis. However, nearly all the selection channels are constructed in a hand-crafted manner, even when they are incorporated into end-to-end deep steganalytic networks, for which the embedding rate and steganographic algorithms also need to be predetermined. Such prior knowledge is usually assumed completely known in existing literature, which is obviously unreasonable and impractical. To address this issue, we propose to automatically learn the selection channels for deep learning-based image steganalysis in a progressive way. Specifically, we divide the image steganalysis task into two phases: selection channel estimation and steganalytic detection. For the first phase, we design a multistage progressive network, which enables the learning of selection channels in a coarse-to-fine fashion. For the second phase, we integrate the learned selection channels into the multilayers of the steganalytic network, allowing full exploitation of selection channels for accurate detection. Extensive experiments demonstrate that the proposed method can learn the selection channels rapidly and precisely, and also significantly improve the detection accuracy of the existing state-of-the-art steganographic network without any prior knowledge.
Open Research
DATA AVAILABILITY STATEMENT
We declare that the data that support the findings of this study are openly available in [BOSSBase] http://agents.fel.cvut.cz/boss/index.php?mode=VIEW%26tmpl=about and [BOWS-2] http://bows2.ec-lille.fr/index.php?mode=VIEW%26tmpl=index1. Data sharing is not applicable to this article as no new data were created or analyzed in this study.
REFERENCES
- 1Li B, He J, Huang J, Shi YQ. A survey on image steganography and steganalysis. J Inf Hiding Multimedia Signal Process. 2011; 2(2): 142-172.
- 2Kadhim IJ, Premaratne P, Vial PJ, Halloran B. Comprehensive survey of image steganography: techniques, evaluations, and trends in future research. Neurocomputing. 2019; 335: 299-326.
- 3Bernard S, Bas P, Klein J, Pevný T. Explicit optimization of min max steganographic game. IEEE Trans Inf Forensics Secur. 2021; 16: 812-823.
- 4Jia J, Luo M, Liu J, Ren W, Wang L. Multiperspective progressive structure adaptation for JPEG steganography detection across domains. IEEE Trans Neural Networks Learn Syst. 2021. doi:10.1109/TNNLS.2021.3054045
- 5Chaumont M. Deep learning in steganography and steganalysis. In: Digital Media Steganography. Elsevier; 2020: 321-349.
10.1016/B978-0-12-819438-6.00022-0 Google Scholar
- 6Jia J, Zhai L, Ren W, Wang L, Ren Y. An effective imbalanced JPEG steganalysis scheme based on adaptive cost-sensitive feature learning. IEEE Trans Knowl Data Eng. 2022; 34(3): 1038-1052.
- 7Wu T, Ren W, Li D, Wang L, Jia J. JPEG steganalysis based on denoising network and attention module. Int J Intell Syst. 2021. doi:10.1002/int.22749
- 8Jia J, Zhai L, Ren W, Wang L, Ren Y, Zhang L. Transferable heterogeneous feature subspace learning for JPEG mismatched steganalysis. Pattern Recognit. 2020; 100:107105.
- 9Holub V, Fridrich JJ. Designing steganographic distortion using directional filters. In: P Moulin, F Pérez-González, eds. 2012 IEEE International Workshop on Information Forensics and Security, WIFS 2012, December 2-5, 2012. IEEE; 2012: 234-239.
10.1109/WIFS.2012.6412655 Google Scholar
- 10Holub V, Fridrich JJ. Digital image steganography using universal distortion. In: W Puech, M Chaumont, J Dittmann, P Campisi, eds. ACM Information Hiding and Multimedia Security Workshop, IH&MMSec'13, June 17-19, 2013. ACM; 2013: 59-68.
10.1145/2482513.2482514 Google Scholar
- 11Xie G, Ren J, Marshall S, Zhao H, Li H. A new cost function for spatial image steganography based on 2D-SSA and WMF. IEEE Access. 2021; 9: 30604-30614.
- 12Zhou W, Zhang W, Yu N. A new rule for cost reassignment in adaptive steganography. IEEE Trans Inf Forensics Secur. 2017; 12(11): 2654-2667.
- 13Yang J, Ruan D, Huang J, Kang X, Shi YQ. An embedding cost learning framework using gan. IEEE Trans Inf Forensics Secur. 2019; 15: 839-851.
- 14Tang W, Li B, Barni M, Li J, Huang J. An automatic cost learning framework for image steganography using deep reinforcement learning. IEEE Trans Inf Forensics Secur. 2020; 16: 952-967.
- 15Tang W, Li H, Luo W, Huang J. Adaptive steganalysis against WOW embedding algorithm. In: A Unterweger, A Uhl, S Katzenbeisser, R Kwitt, A Piva, eds. Proceedings of the 2nd ACM Workshop on Information Hiding and Multimedia Security. ACM; 2014: 91-96.
- 16Denemark T, Sedighi V, Holub V, Cogranne R, Fridrich J. Selection-channel-aware rich model for steganalysis of digital images. In: Y Sun, HV Zhao, eds. 2014 IEEE International Workshop on Information Forensics and Security (WIFS). IEEE; 2014: 48-53.
- 17Tang W, Li H, Luo W, Huang J. Adaptive steganalysis based on embedding probabilities of pixels. IEEE Trans Inf Forensics Secur. 2015; 11(4): 734-745.
- 18Denemark T, Fridrich J, Comesaña-Alfaro P. Improving selection-channel-aware steganalysis features. Electron Imaging. 2016; 2016(8): 1-8.
10.2352/ISSN.2470-1173.2016.8.MWSF-080 Google Scholar
- 19Holub V, Fridrich J. Low-complexity features for JPEG steganalysis using undecimated DCT. IEEE Trans Inf Forensics Secur. 2014; 10(2): 219-228.
- 20Song X, Liu F, Yang C, Luo X, Zhang Y. Steganalysis of adaptive JPEG steganography using 2D Gabor filters. In: AM Alattar, J Fridrich, N Smith, P Comesana-Alfaro, eds. Proceedings of the 3rd ACM Workshop on Information Hiding and Multimedia Security. ACM; 2015: 15-23.
- 21Cogranne R. Selection-channel-aware reverse JPEG compatibility for highly reliable steganalysis of JPEG images. In: A Perez-Neira, X Mestre, eds. 2020 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP 2020, May 4-8, 2020. IEEE; 2020: 2772-2776.
10.1109/ICASSP40776.2020.9054061 Google Scholar
- 22Zhou S, Tang W, Tan S, Li B. Content-adaptive steganalysis via augmented utilization of selection-channel information. In: CD Yoo, YQ Shi, H Kim, A Piva, G Kim, eds. Digital Forensics and Watermarking—Proceedings of the 17th International Workshop, IWDW 2018, October 22-24, 2018, Lecture Notes in Computer Science. Vol. 11378. Springer; 2018: 261-274.
- 23Xu X, Song X, Zhao R. Selection-channel-aware scale co-occurrence matrix feature to detect content-adaptive JPEG steganography. In: Y Tan, Z Ming, eds. Proceedings of the 2018 2nd International Conference on Computer Science and Artificial Intelligence, CSAI 2018, the 10th International Conference on Information and Multimedia Technology, ICIMT 2018, December 8-10, 2018. ACM; 2018: 375-379.
10.1145/3297156.3297212 Google Scholar
- 24Yang J, Liu K, Kang X, Wong EK, Shi Y. Steganalysis based on awareness of selection-channel and deep learning. In: C Kraetzer, YQ Shi, J Dittmann, H Kim, eds. Digital Forensics and Watermarking—Proceedings of the 16th International Workshop, IWDW 2017, August 23-25, 2017. Vol. 10431. Springer; 2017: 263-272.
10.1007/978-3-319-64185-0_20 Google Scholar
- 25Ye J, Ni J, Yi Y. Deep learning hierarchical representations for image steganalysis. IEEE Trans Inf Forensics Secur. 2017; 12(11): 2545-2557.
- 26Boroumand M, Chen M, Fridrich J. Deep residual network for steganalysis of digital images. IEEE Trans Inf Forensics Secur. 2018; 14(5): 1181-1193.
- 27You W, Zhang H, Zhao X. A siamese CNN for image steganalysis. IEEE Trans Inf Forensics Secur. 2021; 16: 291-306.
- 28Zhang R, Zhu F, Liu J, Liu G. Depth-wise separable convolutions and multi-level pooling for an efficient spatial CNN-based steganalysis. IEEE Trans Inf Forensics Secur. 2020; 15: 1138-1150.
- 29Fridrich JJ, Kodovský J. Rich models for steganalysis of digital images. IEEE Trans Inf Forensics Secur. 2012; 7(3): 868-882.
- 30Tang W, Li H, Luo W, Huang J. Adaptive steganalysis against WOW embedding algorithm. In: A Unterweger, A Uhl, S Katzenbeisser, R Kwitt, A Piva, eds. ACM Information Hiding and Multimedia Security Workshop, IH&MMSec'14, June 11-13, 2014. ACM; 2014: 91-96.
10.1145/2600918.2600935 Google Scholar
- 31Denemark T, Sedighi V, Holub V, Cogranne R, Fridrich JJ. Selection-channel-aware rich model for steganalysis of digital images. In: Y Sun, HV Zhao, eds. 2014 IEEE International Workshop on Information Forensics and Security, WIFS 2014, December 3-5, 2014. IEEE; 2014: 48-53.
10.1109/WIFS.2014.7084302 Google Scholar
- 32Denemark T, Fridrich JJ, Alfaro PC. Improving selection-channel-aware steganalysis features. In: AM Alattar, ND Memon, eds. Media Watermarking, Security, and Forensics 2016, February 14-18, 2016. Ingenta; 2016: 1-8.
10.2352/ISSN.2470-1173.2016.8.MWSF-080 Google Scholar
- 33Tan S, Wu W, Shao Z, Li Q, Li B, Huang J. CALPA-NET: channel-pruning-assisted deep residual network for steganalysis of digital images. IEEE Trans Inf Forensics Secur. 2021; 16: 131-146.
- 34Ai S, Hong S, Zheng X, Wang Y, Liu X. CSRT rumor spreading model based on complex network. Int J Intell Syst. 2021; 36(5): 1903-1913.
- 35Ren W, Zhai L, Jia J, Wang L, Zhang L. Learning selection channels for image steganalysis in spatial domain. Neurocomputing. 2020; 401: 78-90.
- 36Mo K, Huang T, Xiang X. Querying little is enough: model inversion attack via latent information. In: XF Chen, HY Yan, QB Yan, XL Zhang, eds. International Conference on Machine Learning for Cyber Security. Springer; 2020: 583-591.
10.1007/978-3-030-62460-6_52 Google Scholar
- 37Kim J, Lee JK, Lee KM. Deeply-recursive convolutional network for image super-resolution. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. IEEE; 2016: 1637-1645.
- 38Wu C, Li W. Enhancing intrusion detection with feature selection and neural network. Int J Intell Syst. 2021; 36(7): 3087-3105.
- 39Shi Z, Chang C, Chen H, Du X, Zhang H. PR-NET: progressively-refined neural network for image manipulation localization. Int J Intell Syst. 2022 . doi:10.1002/int.22822
10.1002/int.22822 Google Scholar
- 40Liu J, Zhu K, Lu W, Luo X, Zhao X. A lightweight 3D convolutional neural network for deepfake detection. Int J Intell Syst. 2021; 36(9): 4990-5004.
- 41Hochreiter S, Schmidhuber J. Long short-term memory. Neural Comput. 1997; 9(8): 1735-1780.
- 42Bas P, Filler T, Pevnỳ T. “Break our steganographic system”: the ins and outs of organizing boss. In: T Filler, T Pevný, S Craver, AD Ker, eds. International Workshop on Information Hiding. Springer; 2011: 59-70.
10.1007/978-3-642-24178-9_5 Google Scholar
- 43Jia J., Luo M., Ma S., Wang L., & Liu, Y. Consensus-clustering-based automatic distribution matching for cross-domain image steganalysis. IEEE Transactions on Knowledge and Data Engineering. 2022. doi:10.1109/tkde.2022.3155924
