ABSTRACT
While deep learning has revolutionized image steganalysis in terms of performance, little is known about how much modern data driven detectors can still be improved. In this paper, we approach this difficult and currently wide open question by working with artificial but realistic looking images with a known statistical model that allows us to compute the detectability of modern content-adaptive algorithms with respect to the most powerful detectors. Multiple artificial image datasets are crafted with different levels of content complexity and noise power to assess their influence on the gap between both types of detectors. Experiments with SRNet as the heuristic detector indicate that independent noise contributes less to the performance gap than content of the same MSE. While this loss is rather small for smooth images, it can be quite large for textured images. A network trained on many realizations of a fixed textured scene will, however, recuperate most of the loss, suggesting that networks have the capacity to approximately learn the parameters of a cover source narrowed to a fixed scene.
- P. Bas. An embedding mechanism for natural steganography after down-sampling. In IEEE ICASSP, New Orleans, LA, March 5--9, 2017.Google ScholarDigital Library
- P. Bas, T. Filler, and T. Pevný. Break our steganographic system -- the ins and outs of organizing BOSS. In T. Filler, T. Pevný, A. Ker, and S. Craver, editors, Information Hiding, 13th International Conference, volume 6958 of Lecture Notes in Computer Science, pages 59--70, Prague, Czech Republic, May 18--20, 2011.Google ScholarCross Ref
- S. Bernard, P. Bas, T. Pevný, and J. Klein. Optimizing additive approximations of non-additive distortion functions. In D. Borghys and P. Bas, editors, The 9th ACM Workshop on Information Hiding and Multimedia Security, pages 105--112, Brussels, Belgium, June 22--25, 2021.Google Scholar
- R. Böhme. Improved Statistical Steganalysis Using Models of Heterogeneous Cover Signals. PhD thesis, Faculty of Computer Science, Technische Universität Dresden, Germany, 2008.Google Scholar
- M. Boroumand, M. Chen, and J. Fridrich. Deep residual network for steganalysis of digital images. IEEE Transactions on Information Forensics and Security, 14(5):1181-- 1193, May 2019.Google ScholarDigital Library
- M. Boroumand, J. Fridrich, and R. Cogranne. Are we there yet? In A. Alattar and N. D. Memon, editors, Proceedings IS&T, Electronic Imaging, Media Watermarking, Security, and Forensics 2019, San Francisco, CA, January 14--17, 2019.Google Scholar
- J. Butora, Y. Yousfi, and J. Fridrich. How to pretrain for steganalysis. In D. Borghys and P. Bas, editors, The 9th ACM Workshop on Information Hiding and Multimedia Security, Brussels, Belgium, 2021. ACM Press.Google Scholar
- K. Chubachi. An ensemble model using CNNs on different domains for ALASKA2 image steganalysis. In IEEE International Workshop on Information Forensics and Security, New York, NY, December 6--11, 2020.Google ScholarCross Ref
- R. Cogranne and J. Fridrich. Modeling and extending the ensemble classifier for steganalysis of digital images using hypothesis testing theory. IEEE Transactions on Information Forensics and Security, 10(2):2627--2642, December 2015.Google ScholarDigital Library
- R. Cogranne, Q. Giboulot, and P. Bas. The ALASKA steganalysis challenge: A first step towards steganalysis "Into the wild". In R. Cogranne and L. Verdoliva, editors, The 7th ACM Workshop on Information Hiding and Multimedia Security, Paris, France, July 3--5, 2019. ACM Press.Google ScholarDigital Library
- R. Cogranne, Q. Giboulot, and P. Bas. ALASKA--2: Challenging academic re- search on steganalysis with realistic images. In IEEE International Workshop on Information Forensics and Security, New York, NY, December 6--11, 2020.Google ScholarCross Ref
- R. Cogranne, V. Sedighi, T. Pevný, and J. Fridrich. Is ensemble classifier needed for steganalysis in high-dimensional feature spaces? In IEEE International Workshop on Information Forensics and Security, Rome, Italy, November 16--19, 2015.Google ScholarCross Ref
- T. Denemark, P. Bas, and J. Fridrich. Natural steganography in JPEG compressed images. In A. Alattar and N. D. Memon, editors, Proceedings IS&T, Electronic Imaging, Media Watermarking, Security, and Forensics 2018, San Francisco, CA, January 29--February 1, 2018.Google Scholar
- T. Denemark, M. Boroumand, and J. Fridrich. Steganalysis features for content- adaptive JPEG steganography. IEEE Transactions on Information Forensics and Security, 11(8):1736--1746, August 2016.Google ScholarDigital Library
- T. Denemark, V. Sedighi, V. Holub, R. Cogranne, and J. Fridrich. Selection- channel-aware rich model for steganalysis of digital images. In IEEE International Workshop on Information Forensics and Security, Atlanta, GA, December 3--5, 2014. 16] E. Dworetzky, E. Kaziakhmedov, and J. Fridrich. On comparing ad hoc detectors with statistical hypothesis tests. In Y. Yousfi, A. Bharati, and C. Pasquini, editors, The 11th ACM Workshop on Information Hiding and Multimedia Security, Chicago, IL, June 28--30, 2023. ACM Press.Google ScholarCross Ref
- A. Foi, M. Trimeche, V. Katkovnik, and K. Egiazarian. Practical poissonian- gaussian noise modeling and fitting for single-image raw-data. IEEE Transactions on Image Processing, 17(10):1737--1754, Oct. 2008.Google ScholarDigital Library
- J. Fridrich and J. Kodovský. Rich models for steganalysis of digital images. IEEE Transactions on Information Forensics and Security, 7(3):868--882, June 2011.Google ScholarDigital Library
- G. E. Healey and R. Kondepudy. Radiometric CCD camera calibration and noise estimation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 16(3):267--276, March 1994.Google ScholarDigital Library
- V. Holub and J. Fridrich. Low-complexity features for JPEG steganalysis using undecimated DCT. IEEE Transactions on Information Forensics and Security, 10(2):219--228, February 2015.Google ScholarDigital Library
- D. Hu, L. Wang, W. Jiang, S. Zheng, and B. Li. A novel image steganography method via deep convolutional generative adversarial networks. IEEE Access, 6:38303--38314, 2018.Google ScholarCross Ref
- J. Huang, J. Ni, L. Wan, and J. Yan. A customized convolutional neural network with low model complexity for JPEG steganalysis. In R. Cogranne and L. Verdoliva, editors, The 7th ACM Workshop on Information Hiding and Multimedia Security, Paris, France, July 3--5, 2019. ACM Press.Google ScholarDigital Library
- J. R. Janesick. Scientific Charge-Coupled Devices, volume Monograph PM83. Washington, DC: SPIE Press - The International Society for Optical Engineering, January 2001.Google ScholarCross Ref
- J. Kodovský, J. Fridrich, and V. Holub. Ensemble classifiers for steganalysis of digital media. IEEE Transactions on Information Forensics and Security, 7(2):432-- 444, April 2012.Google ScholarDigital Library
- B. Li, M. Wang, and J. Huang. A new cost function for spatial image steganography. In Proceedings IEEE, International Conference on Image Processing, ICIP, Paris, France, October 27--30, 2014.Google ScholarCross Ref
- B. Li, W. Wei, A. Ferreira, and S. Tan. ReST-Net: Diverse activation modules and parallel subnets-based CNN for spatial image steganalysis. IEEE Signal Processing Letters, 25(5):650--654, May 2018.Google ScholarCross Ref
- L. Li, W. Zhang, C. Qin, K. Chen, W. Zhou, and N. Yu. Adversarial batch im- age steganography against CNN-based pooled steganalysis. Signal Processing, 181:107920--107920, 2021.Google ScholarCross Ref
- Ce Liu, R. Szeliski, Sing Bing Kang, C.L. Zitnick, and W.T. Freeman. Automatic estimation and removal of noise from a single image. Pattern Analysis and Machine Intelligence, IEEE Transactions on, 30(2):299--314, Feb 2008.Google ScholarDigital Library
- M. K. Mihcak, I. Kozintsev, K. Ramchandran, and P. Moulin. Low-complexity image denoising based on statistical modeling of wavelet coefficients. IEEE Signal Processing Letters, 6(12):300--303, December 1999.Google ScholarCross Ref
- Y. Qian, J. Dong, W. Wang, and T. Tan. Deep learning for steganalysis via convolu- tional neural networks. In A. Alattar and N. D. Memon, editors, Proceedings SPIE, Electronic Imaging, Media Watermarking, Security, and Forensics 2015, volume 9409, San Francisco, CA, February 8--12, 2015.Google Scholar
- V. Sedighi, R. Cogranne, and J. Fridrich. Content-adaptive steganography by minimizing statistical detectability. IEEE Transactions on Information Forensics and Security, 11(2):221--234, 2016.Google ScholarDigital Library
- Y. Q. Shi, P. Sutthiwan, and L. Chen. Textural features for steganalysis. In M. Kirchner and D. Ghosal, editors, Information Hiding, 14th International Confer- ence, volume 7692 of Lecture Notes in Computer Science, pages 63--77, Berkeley, California, May 15--18, 2012.Google ScholarDigital Library
- X. Song, F. Liu, C. Yang, X. Luo, and Y. Zhang. Steganalysis of adaptive JPEG steganography using 2D Gabor filters. In P. Comesana, J. Fridrich, and A. Alattar, editors, 3rd ACM IH&MMSec. Workshop, Portland, Oregon, June 17--19, 2015.Google Scholar
- W. Su, J. Ni, X. Hu, and J. Huang. New design paradigm of distortion cost function for efficient JPEG steganography. Signal Processing, 190:108319, 2022.Google ScholarCross Ref
- T. Taburet, P. Bas, J. Fridrich, and W. Sawaya. Computing dependencies between DCT coefficients for natural steganography in JPEG domain. In R. Cogranne and L. Verdoliva, editors, The 7th ACM Workshop on Information Hiding and Multimedia Security, Paris, France, July 3--5, 2019. ACM Press.Google ScholarDigital Library
- T. Taburet, P. Bas, W. Sawaya, and J. Fridrich. A natural steganography embed- ding scheme dedicated to color sensors in the JPEG domain. In A. Alattar and N. D. Memon, editors, Proceedings IS&T, Electronic Imaging, Media Watermarking, Security, and Forensics 2019, San Francisco, CA, January 14--17, 2019.Google Scholar
- T. Taburet, P. Bas, W. Sawaya, and J. Fridrich. Natural steganography in JPEG domain with a linear development pipeline. IEEE Transactions on Information Forensics and Security, 16:173--186, 2021.Google ScholarCross Ref
- W. Tang, B. Li, M. Barni, J. Li, and J. Huang. Improving cost learning for JPEG steganography by exploiting JPEG domain knowledge. IEEE Transactions on Circuits and Systems for Video Technology, 32(6):4081--4095, 2021.Google ScholarCross Ref
- W. Tang, B. Li, S. Tan, M. Barni, and J. Huang. CNN-based adversarial embedding for image steganography. IEEE Transactions on Information Forensics and Security, 14(8):2074--2087, 2019.Google ScholarDigital Library
- W. Tang, H. Li, W. Luo, and J. Huang. Adaptive steganalysis based on embedding probabilities of pixels. IEEE Transactions on Information Forensics and Security, 11(4):734--745, April 2016.Google ScholarDigital Library
- Thanh Hai Thai, R. Cogranne, and F. Retraint. Camera model identification based on the heteroscedastic noise model. Image Processing, IEEE Transactions on, 23(1):250--263, Jan 2014.Google ScholarDigital Library
- G. Xu, H. Z. Wu, and Y. Q. Shi. Structural design of convolutional neural networks for steganalysis. IEEE Signal Processing Letters, 23(5):708--712, May 2016.Google ScholarCross Ref
- J. Yang, K. Liu, X. Kang, E. K. Wong, and Y.-Q. Shi. Spatial image steganography based on generative adversarial network, 2018.Google Scholar
- J. Yang, D. Ruan, J. Huang, X. Kang, and Y. Q. Shi. An embedding cost learning framework using GAN. IEEE Transactions on Information Forensics and Security, 15(10):839--851, 2020.Google ScholarDigital Library
- J. Yang, D. Ruan, X. Kang, and Y.-Q. Shi. Towards automatic embedding cost learning for JPEG steganography. In R. Cogranne and L. Verdoliva, editors, The 7th ACM Workshop on Information Hiding and Multimedia Security, Paris, France, July 3--5, 2019. ACM Press.Google ScholarDigital Library
- J. Yang, Y.-Q. Shi, E.K. Wong, and X. Kang. JPEG steganalysis based on densenet. CoRR, abs/1711.09335, 2017.Google Scholar
- J. Ye, J. Ni, and Y. Yi. Deep learning hierarchical representations for image steganalysis. IEEE Transactions on Information Forensics and Security, 12(11):2545-- 2557, November 2017.Google ScholarDigital Library
- M. Yedroudj, F. Comby, and M. Chaumont. Yedroudj-net: An efficient CNN for spatial steganalysis. In IEEE ICASSP, pages 2092--2096, Alberta, Canada, April 15--20, 2018.Google ScholarDigital Library
- Y. Yousfi, J. Butora, and J. Fridrich. CNN steganalyzers leverage local embedding artifacts. In IEEE International Workshop on Information Forensics and Security, Montpellier, France, December 7--10, 2021.Google ScholarCross Ref
- Y. Yousfi, J. Butora, Q. Giboulot, and J. Fridrich. Breaking ALASKA: Color sepa- ration for steganalysis in JPEG domain. In R. Cogranne and L. Verdoliva, editors, The 7th ACM Workshop on Information Hiding and Multimedia Security, Paris, France, July 3--5, 2019. ACM Press.Google ScholarDigital Library
- Y. Yousfi, J. Butora, E. Khvedchenya, and J. Fridrich. ImageNet pre-trained CNNs for JPEG steganalysis. In IEEE International Workshop on Information Forensics and Security, New York, NY, December 6--11, 2020.Google ScholarCross Ref
- Y. Yousfi and J. Fridrich. An intriguing struggle of CNNs in JPEG steganalysis and the OneHot solution. IEEE Signal Processing Letters, 27:830--834, 2020.Google ScholarCross Ref
- J. Zeng, S. Tan, B. Li, and J. Huang. Large-scale JPEG image steganalysis using hybrid deep-learning framework. IEEE Transactions on Information Forensics and Security, 13(5):1200--1214, May 2018.Google ScholarDigital Library
- J. Zeng, S. Tan, G. Liu, Bin Li, and J. Huang. WISERNet: Wider separate-then- reunion network for steganalysis of color images. CoRR, abs/1803.04805, 2018.Google Scholar
- N. Zhong, Z. Qian, Z. Wang, X. Zhang, and X. Li. Batch steganography via gen- erative network. IEEE Transactions on Circuits and Systems for Video Technology, 31(1):88--97, January 2021.Google ScholarDigital Library
Index Terms
- Limits of Data Driven Steganography Detectors
Recommendations
Implementation of Least Significant Bit Steganography and statistical steganalysis
CCSEIT '12: Proceedings of the Second International Conference on Computational Science, Engineering and Information TechnologyLeast Significant Bit (LSB) Steganography is a method for hidden information in such a way that can only be detected by its intended recipient. Steganography in Digital media becomes a challenging discipline, since the Human Auditory System (HAS) is ...
A JPEG-based statistically invisible steganography
ICIMCS '11: Proceedings of the Third International Conference on Internet Multimedia Computing and ServiceThe objective of steganography is to implement covert communication by hiding data in digital files such that there is no indication of the existence of hidden data; so an ideal steganographic system should be able to withstand all steganalysis methods--...
Secure steganography based on embedding capacity
Mostly the embedding capacity of steganography methods is assessed in non-zero DCT coefficients. Due to unequal distribution of non-zero DCT coefficients in images with different contents, images with the same number of non-zero DCT coefficients may ...
Comments