Abstract
Steganalysis, as the opposite technique to steganography, has been applied to determine whether secret information is embedded in an image. The existing adaptive steganography methods embed secret information in different regions of an image with different probabilities. However, it is difficult for most steganalysis models to make a targeted attention of steganography regions of images, which reduces the detection accuracy of the steganalysis models when detecting the adaptive steganography methods. Aiming to detect the JPEG-based adaptive steganography, this paper proposes a model called Steganalysis Attention Augmented Network (SAANet). The proposed model uses attention augmented convolution instead of traditional convolution, so that the network assigns more learning weights to the steganographic area, and guides the network to better learn features that are beneficial to steganalysis. Thereby improving the learning ability and training effect of the model. As far as we know, we are the first to directly apply attention to the backbone structure of the network model to assist in image steganalysis. And our work improves the accuracy and operating efficiency of the steganalysis model. Experiments show that compared with the current steganalysis models, the model proposed in this paper obtains a competitive detection performance. The current adaptive steganography algorithm achieves a detection accuracy of up to 96.68%, while 95.22% under mismatch conditions, and the testing time is 28 s, indicating that the proposed model has a certain generalization performance and practicability.
Similar content being viewed by others
References
Bas P, Furon T (2007) BOWS-2. http://bows2.ec-lille.fr Accessed 10 Jul 2007
Bas P, Filler T, Pevny T (2011) Break our steganographic system: the ins and outs of organizing BOSS. In: Proceedings of 13th International Conference on Information Hiding, IH’2011. LNCS. Springer, Heidelberg, pp 59–70
Bello I, Zoph B, Le Q, Vaswani A, Shlens J (2019) Attention augmented convolutional networks. arxiv:1904.09925
Boroumand M, Chen M, Fridrich J (2019) Deep residual network for steganalysis of digital images. IEEE Trans Inf Forensic Secur 14(5):1181–1193
Chen M, Sedighi V, Boroumand M, Fridrich J (2017) Jpeg-phase-aware convolutional neural network for steganalysis of JPEG images. In: Proceedings of the 5th ACM workshop on information hiding and multimedia security, IH&MMSec 2017, Philadelphia, June 20–22, 2017, pp 75–84
Cheng J, Dong L, Lapata M (2016) Long short-term memory-networks for machine reading. In: Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing. ACL Press, Stroudsburg, pp 551–561
Dalal M, Juneja M (2020) Steganography and Steganalysis (in digital forensics): a cybersecurity guide. Multimed Tools Appl 80:5723–5771
Denemark T, Sedighi V , Holub V , Cogranne R, Fridrich J (2014) Selection-channel-aware rich model for steganalysis of digital images. In: 2014 IEEE International Workshop on Information forensics and security (WIFS), pp 48–53
Denemark T, Boroumand M, Fridrich J (2016) Steganalysis features for content-adaptive JPEG steganography. IEEE Trans Inf Forensic Secur 11(8):1736–1746
Fridrich J, Kodovsky J (2012) Rich models for steganalysis of digital images. IEEE Trans Inf Forensic Secur 7(3):868–882
Guo L, Ni J, Shi YQ (2014) Uniform embedding for efficient JPEG steganography. IEEE Trans Inf Forensic Secur 9(5):814–825
Guo L, Ni J, Su W (2015) Using statistical image model for JPEG steganography: uniform embedding revisited. IEEE Trans Inf Forensic Secur 10(12):2669–2680
Guttikonda J, Sridevi R (2019) A new steganalysis approach with an efficient feature selection and classification algorithms for identifying the stego images. Multimed Tools Appl 78(15):21113–21131
He K, Zhang X, Ren S, Sun J (2016) Deep residual learning for image recognition. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition
Holub V, Fridrich J (2012) Designing steganographic distortion using directional filters. In: Fourth IEEE International Workshop on Information Forensics and Security
Holub V, Fridrich J (2015) Low-complexity features for JPEG steganalysis using undecimated DCT. IEEE Trans Inf Forensic Secur 10:219–228
Holub V, Fridrich J, Denemark T (2014) Universal distortion function for stegangography in an arbitrary domain. EURASIP J Inf Secur (1):1–13
Hu D, Shen Q, Zhou S, Liu X, Fan Y, Wang L (2017) Adaptive steganalysis based on selection region and combined convolutional neural networks. Sec Commun Netw 2017(4):1–9
Hu D, Zhou S, Shen Q, Zheng S, Zhao Z, Fan Y (2019) Digital image steganalysis based on visual attention and deep reinforcement learning. IEEE Access 7:25924–25935
Jung K-H (2010) High-capacity steganographic method based on pixel-value differencing and LSB replacement methods. Imaging Sci J 58(4):213–221
Jung K-H (2018) Comparative histogram analysis of LSB-based image steganography. WSEAS Trans Syst Control 13:103–112
Jung K-H, Seok W-J (2015) Research on steganography in emulab testbed. J Inst Electron Inf Eng 52(11):79–84
Kaur G, Singh S, Rani R, Kumar R (2020) A comprehensive study of reversible data hiding (RDH) schemes based on pixel value ordering (PVO). Archives of Computational Methods in Engineering pp 1–52
Kingma D P, Ba J (2014) Adam: a method for stochastic optimization. arxiv:1412.6980
Krizhevsky A, Sutskever I , Hinton G (2012) ImageNet classification with deep convolutional neural networks. NIPS
Li B, Wang M, Huang J, Li X (2014) A new cost function for spatial image steganography. IEEE International Conference on Image Processing (ICIP), pp 4206–4210
Li S, Ye D, Jiang S (2019) Anti-steganalysis for image on convolutional neural networks. Multimed Tools Appl 79:4315–4331
Liao X, Chen G, Yin J (2017) Content adaptive steganalysis for color images. Secur Commun Netw 9:5756–5763
Parmar N, Vaswani A, Uszkoreit J (2018) Image transformer. arxiv:1802.05751
Pevný T, Filler T, Bas P (2010) Using high-dimensional image models to perform highly undetectable steganography. Lect Notes Comput Sci 6387:161–177
Qian Y, Dong J, Wang W, Tan T (2015) Deep learning for steganalysis via convolutional neural networks. Proc SPIE Med Watermark Sec Forensic 94:94090J
Rawat R, Singh B, Sur A (2020) Steganalysis for clustering modification directions steganography. Multimed Tools Appl 79(1):1971–1986
Sedighi V, Cogranne R, Fridrich J (2015) Content-adaptive steganography by minimizing statistical detectability. IEEE Trans Inf Forensic Secur 11(2):221–234
Singh B, Chhajed M, Sur A (2020) Steganalysis using learned denoising kernels. Multimed Tools Appl 80:4903–4917
Su A, He X, Zhao X (2020) JPEG steganalysis based on ResNeXt with gauss partial derivative filters. Multimed Tools Appl 1–18
Su A, He X, Zhao X (2021) JPEG steganalysis based on ResNeXt with gauss partial derivative filters. Multimed Tools Appl 80(3):3349–3366
Tang W, Li H, Luo W, Huang J (2016) Adaptive steganalysis based on embedding probabilities of pixels. IEEE Trans Inf Forensic Secur 11(4):734–745
Vaswani A, Shazeer N, Parmar N (2017) Attention is all you need. arxiv:1706.03762
Wang X, Girshick R, Gupta A (2017) Non-local neural networks. arxiv:1711.07971
Xu G (2017) Deep convolutional neural network to detect J-UNIW ARD. In: Proceedings of the 5th ACM workshop on information hiding and multimedia security, IH&MMSec 2017, Philadelphia, June 20–22, 2017, pp 67–73
Xu G, Wu H (2016) Structural design of convolutional neural networks for steganalysis. IEEE Signal Process Lett 23(5):708–712
Ye J, Ni J, Yi Y (2017) Deep learning hierarchical representations for image steganalysis. IEEE Trans Inf Forensic Secur 12(11):2545–2557
Zhang H, Goodfellow I, Metaxas D (2018) Self-attention generative adversarial networks. arxiv:1805.08318
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
About this article
Cite this article
Huang, S., Zhang, M., Ke, Y. et al. Image steganalysis based on attention augmented convolution. Multimed Tools Appl 81, 19471–19490 (2022). https://doi.org/10.1007/s11042-021-11862-4
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11042-021-11862-4