A new JPEG image steganalysis technique combining rich model features and convolutional neural networks

Tao Zhang; Hao Zhang; Ran Wang; Yunda Wu; Tao Zhang; Hao Zhang; Ran Wang; Yunda Wu

doi:10.3934/mbe.2019201

Mathematical Biosciences and Engineering

2019, Volume 16, Issue 5: 4069-4081. doi: 10.3934/mbe.2019201

Previous Article Next Article

Research article Special Issues

A new JPEG image steganalysis technique combining rich model features and convolutional neural networks

1.
School of Computer Science and Engineering, Changshu Institute of Technology, No.99, Hushan Road, Changshu 215500, Jiangsu Province, China
2.
National Digital Switching System Engineering and Technology Research Center, No.62, Kexue Avenue, Zhengzhou 450001, Henan Province, China

Received: 21 January 2019 Accepted: 29 April 2019 Published: 07 May 2019

The best traditional steganalysis methods aiming at adaptive steganography are the combination of rich models and ensemble classifier. In this study, a new steganalysis method for JPEG images based on convolutional neural networks is proposed to solve the high dimension problem in steganalysis from another aspect. On the basis of the original rich model, the algorithm adds different sizes of discrete cosine transform (DCT) basis functions to extract different detection features. Different features are combined at the fully connected layer through inputting 2-D feature values to the neural network convolutional layer for predictive classification. Experimental results show that convolutional neural networks as classifiers do not require a large number of training samples, and the final classification performance is better than that of the original ensemble classifier.

Keywords:

Citation: Tao Zhang, Hao Zhang, Ran Wang, Yunda Wu. A new JPEG image steganalysis technique combining rich model features and convolutional neural networks[J]. Mathematical Biosciences and Engineering, 2019, 16(5): 4069-4081. doi: 10.3934/mbe.2019201

Related Papers:

[1]	Verónica Martínez-Fernández, Francisco Navarro . Rpb5, a subunit shared by eukaryotic RNA polymerases, cooperates with prefoldin-like Bud27/URI. AIMS Genetics, 2018, 5(1): 63-74. doi: 10.3934/genet.2018.1.63
[2]	Jue Er Amanda Lee, Linda May Parsons, Leonie M. Quinn . MYC function and regulation in flies: how Drosophila has enlightened MYC cancer biology. AIMS Genetics, 2014, 1(1): 81-98. doi: 10.3934/genet.2014.1.81
[3]	Carlos García-Padilla, Amelia Aránega, Diego Franco . The role of long non-coding RNAs in cardiac development and disease. AIMS Genetics, 2018, 5(2): 124-140. doi: 10.3934/genet.2018.2.124
[4]	Tú Nguyen-Dumont, Jenna Stewart, Ingrid Winship, Melissa C. Southey . Rare genetic variants: making the connection with breast cancer susceptibility. AIMS Genetics, 2015, 2(4): 281-292. doi: 10.3934/genet.2015.4.281
[5]	Valeriia Mikhailova, Valeriia Gulaia, Vladlena Tiasto, Stanislav Rybtsov, Margarita Yatsunskaya, Alexander Kagansky . Towards an advanced cell-based in vitro glioma model system. AIMS Genetics, 2018, 5(2): 91-112. doi: 10.3934/genet.2018.2.91
[6]	Shafagh A. Waters, Paul D. Waters . Imprinted X chromosome inactivation: evolution of mechanisms in distantly related mammals. AIMS Genetics, 2015, 2(2): 110-126. doi: 10.3934/genet.2015.2.110
[7]	Jeffrey M. Marcus . Our love-hate relationship with DNA barcodes, the Y2K problem, and the search for next generation barcodes. AIMS Genetics, 2018, 5(1): 1-23. doi: 10.3934/genet.2018.1.1
[8]	Vladlena Tiasto, Valeriia Mikhailova, Valeriia Gulaia, Valeriia Vikhareva, Boris Zorin, Alexandra Kalitnik, Alexander Kagansky . Esophageal cancer research today and tomorrow: Lessons from algae and other perspectives. AIMS Genetics, 2018, 5(1): 75-90. doi: 10.3934/genet.2018.1.75
[9]	Ke-Sheng Wang, Xuefeng Liu, Daniel Owusu, Yue Pan, Changchun Xie . Polymorphisms in the ANKS1B gene are associated with cancer, obesity and type 2 diabetes. AIMS Genetics, 2015, 2(3): 192-203. doi: 10.3934/genet.2015.3.192
[10]	Asaad M Mahmood, Jim M Dunwell . Evidence for novel epigenetic marks within plants. AIMS Genetics, 2019, 6(4): 70-87. doi: 10.3934/genet.2019.4.70

Abstract

Colon cancer is one of the leading causes of mortality and morbidity worldwide [1], with a dramatic rise in incidence and mortality rates over the past 3 decades, both worldwide and in China [2]. Surgery is the principal therapy for early-stage colon cancer because it offers the only method for complete tumor removal, thereby chance for cure [3]. However, there is still proportions of colon cancer patients that develop recurrence even after radical resection [4]. Unfortunately, long term control of the recurrent colon cancer has been a difficult dilemma to tackle and the clinical outcomes of these patients are poor [5],[6]. Adjuvant chemotherapy in colon cancer is able to complement curative surgery to reduce the risk of recurrence and death from relapsed or metastatic disease [7],[8]. However, patient willingness to be adjuvantly treated has been limited due to its associated financial burden, treatment toxicity and debatable prolongation of survival that need to be carefully balanced against the associated-toxicity. Therefore, the identification of robust biomarkers to assess the risk of postoperative recurrence in colon cancer has been a longing need.

Circular RNAs (circRNAs), a recently discovered type of noncoding RNA, have a special circular structure formed by 3′- and 5′-ends linking covalently [9]. Because they do not have 5′ or 3′ ends, they are resistant to exonuclease-mediated degradation [10]. Therefore, circRNAs are characterized by their high stability, expression level, and evolutionary conservation. Recent literatures [11]–[13] have confirmed circRNAs as playing an important role in cancer initiation and progression. Additionally, they are differentially expressed in cancer tissues and circulation of cancer patients. Due to these characteristics, circRNAs have gained recognition as a promising novel biomarker for cancer [14]. However, till now little is known about circRNAs and their relationship with colon cancer. In a recent study published in EMBO Molecular Medicine, entitled “A circRNA signature predicts postoperative recurrence in stage II/III colon cancer”, Ju et al. [15] identified and validated a circRNA-based signature that could improve postoperative prognostic stratification of patients with stage II/III colon cancer.

In that study, 437 colon cancer circRNAs were examined in tumoral and adjacent normal tissues. The authors found that 103 circRNAs were differentially expressed in recurrent colon cancer patients as compared to non-recurrent counterparts. Among them, 100 up-regulated circRNAs were validated in patients with stage II/III colon cancer to test whether they could be used as prognostic biomarkers. Four circRNAs, including hsa_circ_0122319, hsa_circ_0087391, hsa_circ_0079480, and hsa_circ_0008039, showed strong predictive values for disease-free survival (DFS) in the validation cohort. The following circRNA-based prognostic model was thereby generated:

c i r S c o r e = 0. 46 \times {Exp}_{h s a_c i r c_0122319} + (- 0. 386 \times {Exp}_{h s a_c i r c_0083791}) + 0. 293 \times {Exp}_{h s a_c i r c_0079480} + 0. 439 \times {Exp}_{h s a_c i r c_0008039}

$c i r S c o r e = 0. 46 \times {Exp}_{h s a_c i r c_0122319} + (- 0. 386 \times {Exp}_{h s a_c i r c_0083791}) + 0. 293 \times {Exp}_{h s a_c i r c_0079480} + 0. 439 \times {Exp}_{h s a_c i r c_0008039}$

After adjusting for baseline clinicopathological factors, the cirScore was found suitable for serving as predictive biomarkers of postoperative recurrence in stage II/III colon cancer. Patients in the high-risk group (cirScore ≥ −0.323) had poorer DFS (hazard ratio [HR], 2.89, 95% confidence interval [CI], 1.37–6.09, P < 0.001) and overall survival (OS; HR, 4.22, 95% CI, 1.61–11.03, P < 0.001) than those in the low-risk group (cirScore < −0.323). Based on this circRNA-based prognostic model, the authors further established a nomogram which exhibited good performance for estimating the 3- or 5-year DFS and OS.

To further explore the biological functions of hsa_circ_0122319, hsa_circ_0087391, and hsa_circ_0079480 (the abundance of hsa_circ_0008039 was found to be low in colon cancer cell line was therefore not further explored), the authors performed a series of functional in vitro and in vivo experiments. They found that suppression of these three circRNAs in colon cancer cells could significantly alter in vitro migration capacity. Additionally, the knockdown of hsa_circ_0079480 in colon cancer cells could both inhibit in vivo lung and liver metastasis.

The strengths of this study include the well-designed identification approaches, mature clinical data with clinical outcome and long-term follow-up, and well-characterized tissue samples for colon cancer. Limitations included that the prognostic model was only applicable for the patients with stage II/III colon cancer. In summary, this study identified a 4-circRNA-based expression signature as highly predictive for cancer recurrence in stage II/III colon cancer, with the potential of identification of high-risk early-stage colon cancer individuals who would benefit most from adjuvant chemotherapy.

References

[1]	V. Holub and J. Fridrich, Low-complexity features for jpeg steganalysis using undecimated DCT, IEEE Trans. Inf. Fore. Secu., 10(2015), 219–228.
[2]	J. Fridrich and J. Kodovsky, Rich models for steganalysis of digital images, IEEE Trans. Inf. Fore. Secu., 7(2012), 868–882.
[3]	V. Holub and J. Fridrich, Phase-aware projection model for steganalysis of JPEG images, Proc. SPIE, 9409(2015), 1–11.
[4]	J. Kodovsky, J. Fridrich and V. Holub, Ensemble classifiers for steganalysis of digital media, IEEE Trans. Inf. Fore. Secu., 7(2012), 432–444.
[5]	J. Fridrich, Steganalysis in high dimensions: fusing classifiers built on random subspaces, Proc. SPIE, 7880(2011), 181–197.
[6]	V. Holub and J. Fridrich, Random projections of residuals for digital image steganalysis, IEEE Trans. Inf. Fore. Secu., 8(2013), 1996–2006.
[7]	J. Kodovský and J. Fridrich, Steganalysis in high dimensions: fusing classifiers built on random subspaces, Proc. SPIE, 7880(2011), 1–13.
[8]	S. Tan and B. Li, Stacked convolutional auto-encoders for steganalysis of digital images, Proc. IEEE APSIPA, Siem Reap, Cambodia, (2014), 1–4.
[9]	T. Pevný, T. Filler and P. Bas, Using high-dimensional image models to perform highly undetectable steganography, Proc. IHW, (2010), 161–177.
[10]	Y. Qian, J. Dong and W. Wang, Deep learning for steganalysis via convolutional neural networks, Proc. SPIE, 9409(2015), 1–10.
[11]	V. Holub and J. Fridrich, Designing steganographic distortion using directional filters, Proc. IEEE IFS, (2012), 234–239.
[12]	V. Holub, J. Fridrich and T. Denemark, Universal distortion function for steganography in an arbitrary domain, EURASIP J. Info. Sec., 2014(2014), 1–13.
[13]	L. Pibre L, J. Pasquet and D. Ienco, Deep learning is a good steganalysis tool when embedding key is reused for different images, even if there is a cover sourcemismatch, Proc. EI'2016, (2016), 79–95.
[14]	Mo Chen, V. Sedighi, M. Boroumand, et al., JPEG-phase-aware convolutional neural network for steganalysis of JPEG images, Proc. IH MMSec, (2017), 1–10.
[15]	J. Zeng, S. Tan, B. Li, et al., Large-scale JPEG image steganalysis using hybrid deep-learning framework, IEEE Trans. Info. Forens. Secu., 13(2018), 1200–1214.
[16]	G. Xu, Deep convolutional neural network to detect JUNIWARD, Proc. IH MMSec, (2017), 1–6.
[17]	M. Boroumand, M. Chen and J. Fridrich, Deep residual network for steganalysis of digital images. IEEE Trans. Info. Fore. Secu., 14(2019), 1181–1193.
[18]	L. Chang, X. Den and M. Zhou, Convolution neural networks in image understanding, Acta Automat. Sinica, 42(2016), 1300–1312.
[19]	X. Glorot and Y. Bengio, Understanding the difficulty of training deep feedforward neural networks, J. Mach. Lear. Rese., 9(2010), 249–256.
[20]	S. Ioffe and C. Szegedy, Batch normalization: Accelerating deep network training by reducing internal covariate shift, Proc. ICML, (2015), 448–456.
[21]	N. Srivastava, G. Hinton and A. Krizhevsky, Dropout: a simple way to prevent neural networks from overfitting, J. Mach. Lear. Rese., 15(2014), 1929–1958.
[22]	L. Guo, J. Ni and Y. Shi, Uniform embedding for efficient JPEG steganography, IEEE Trans. Info. Fore. Secu., 9(2014), 814–825.
[23]	J. Bergstra, O. Breuleux and F. Bastien, Theano: A CPU and GPU math compiler in Python, Proc. PSC, (2010), 1–7.

This article has been cited by:

Ya’nan Zhen, Guodong Sun, Cunbao Chen, Jianqi Li, Ruixue Xiao, Zhongfa Xu, Circular RNA hsa_circ_0064559 affects tumor cell growth and progression of colorectal cancer, 2023, 21, 1477-7819, 10.1186/s12957-023-03050-5

Reader Comments

Your name:*

Email:*
© 2019 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)