How to augment a small learning set for improving the performances of a CNN-based steganalyzer?

Deep learning and convolutional neural networks (CNN) have been intensively used in many image processing topics during last years. As far as steganalysis is concerned, the use of CNN allows reaching the state-of-the-art results. The performances of such networks often rely on the size of their learning database. An obvious preliminary assumption could be considering that "the bigger a database is, the better the results are". However, it appears that cautions have to be taken when increasing the database size if one desire to improve the classification accuracy i.e. enhance the steganalysis efficiency. To our knowledge, no study has been performed on the enrichment impact of a learning database on the steganalysis performance. What kind of images can be added to the initial learning set? What are the sensitive criteria: the camera models used for acquiring the images, the treatments applied to the images, the cameras proportions in the database, etc? This article continues the work carried out in a previous paper, and explores the ways to improve the performances of CNN. It aims at studying the effects of "base augmentation" on the performance of steganalysis using a CNN. We present the results of this study using various experimental protocols and various databases to define the good practices in base augmentation for steganalysis.

[1]  Jing Dong,et al.  Deep learning for steganalysis via convolutional neural networks , 2015, Electronic Imaging.

[2]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[3]  Jessica J. Fridrich,et al.  Designing steganographic distortion using directional filters , 2012, 2012 IEEE International Workshop on Information Forensics and Security (WIFS).

[4]  Trevor Darrell,et al.  Caffe: Convolutional Architecture for Fast Feature Embedding , 2014, ACM Multimedia.

[5]  Jessica J. Fridrich,et al.  Ensemble Classifiers for Steganalysis of Digital Media , 2012, IEEE Transactions on Information Forensics and Security.

[6]  Jiangqun Ni,et al.  Deep Learning Hierarchical Representations for Image Steganalysis , 2017, IEEE Transactions on Information Forensics and Security.

[7]  Jiwu Huang,et al.  Large-Scale JPEG Image Steganalysis Using Hybrid Deep-Learning Framework , 2016, IEEE Transactions on Information Forensics and Security.

[8]  Mauro Barni,et al.  A Comparative Study of ±1 Steganalyzers , 2008 .

[9]  Yun Q. Shi,et al.  Ensemble of CNNs for Steganalysis: An Empirical Study , 2016, IH&MMSec.

[10]  Marc Chaumont,et al.  Deep learning is a good steganalysis tool when embedding key is reused for different images, even if there is a cover sourcemismatch , 2015, Media Watermarking, Security, and Forensics.

[11]  Marc Chaumont,et al.  Yedroudj-Net: An Efficient CNN for Spatial Steganalysis , 2018, 2018 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP).

[12]  Jessica J. Fridrich,et al.  Rich Models for Steganalysis of Digital Images , 2012, IEEE Transactions on Information Forensics and Security.

[13]  Guanshuo Xu,et al.  Deep Convolutional Neural Network to Detect J-UNIWARD , 2017, IH&MMSec.

[14]  Yun Q. Shi,et al.  Structural Design of Convolutional Neural Networks for Steganalysis , 2016, IEEE Signal Processing Letters.

[15]  Marc Chaumont The emergence of Deep Learning in steganography and steganalysis , 2018 .

[16]  Mo Chen,et al.  JPEG-Phase-Aware Convolutional Neural Network for Steganalysis of JPEG Images , 2017, IH&MMSec.

[17]  Jessica J. Fridrich,et al.  Study of cover source mismatch in steganalysis and ways to mitigate its impact , 2014, Electronic Imaging.

[18]  Tomás Pevný,et al.  "Break Our Steganographic System": The Ins and Outs of Organizing BOSS , 2011, Information Hiding.

[19]  Rainer Böhme,et al.  Moving steganography and steganalysis from the laboratory into the real world , 2013, IH&MMSec '13.

[20]  Jessica J. Fridrich,et al.  Universal distortion function for steganography in an arbitrary domain , 2014, EURASIP Journal on Information Security.

[21]  Jing Dong,et al.  Learning and transferring representations for image steganalysis using convolutional neural network , 2016, 2016 IEEE International Conference on Image Processing (ICIP).

[22]  Bolei Zhou,et al.  Places: A 10 Million Image Database for Scene Recognition , 2018, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[23]  Tomás Pevný,et al.  A mishmash of methods for mitigating the model mismatch mess , 2014, Electronic Imaging.