Adaptive Steganalysis Based on Selection Region and Combined Convolutional Neural Networks
Digital image steganalysis is the art of detecting the presence of information hiding in carrier images. When detecting recently developed adaptive image steganography methods, state-of-art steganalysis methods cannot achieve satisfactory detection accuracy, because the adaptive steganography method...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Hindawi-Wiley
2017-01-01
|
Series: | Security and Communication Networks |
Online Access: | http://dx.doi.org/10.1155/2017/2314860 |
id |
doaj-5682396c3132466381d6eb14a789c4df |
---|---|
record_format |
Article |
spelling |
doaj-5682396c3132466381d6eb14a789c4df2020-11-25T01:11:50ZengHindawi-WileySecurity and Communication Networks1939-01141939-01222017-01-01201710.1155/2017/23148602314860Adaptive Steganalysis Based on Selection Region and Combined Convolutional Neural NetworksDonghui Hu0Qiang Shen1Shengnan Zhou2Xueliang Liu3Yuqi Fan4Lina Wang5School of Computer and Information, Hefei University of Technology, Hefei 230009, ChinaSchool of Computer and Information, Hefei University of Technology, Hefei 230009, ChinaSchool of Computer and Information, Hefei University of Technology, Hefei 230009, ChinaSchool of Computer and Information, Hefei University of Technology, Hefei 230009, ChinaSchool of Computer and Information, Hefei University of Technology, Hefei 230009, ChinaSchool of Computer and Information, Wuhan University, Wuhan 430072, ChinaDigital image steganalysis is the art of detecting the presence of information hiding in carrier images. When detecting recently developed adaptive image steganography methods, state-of-art steganalysis methods cannot achieve satisfactory detection accuracy, because the adaptive steganography methods can adaptively embed information into regions with rich textures via the guidance of distortion function and thus make the effective steganalysis features hard to be extracted. Inspired by the promising success which convolutional neural network (CNN) has achieved in the fields of digital image analysis, increasing researchers are devoted to designing CNN based steganalysis methods. But as for detecting adaptive steganography methods, the results achieved by CNN based methods are still far from expected. In this paper, we propose a hybrid approach by designing a region selection method and a new CNN framework. In order to make the CNN focus on the regions with complex textures, we design a region selection method by finding a region with the maximal sum of the embedding probabilities. To evolve more diverse and effective steganalysis features, we design a new CNN framework consisting of three separate subnets with independent structure and configuration parameters and then merge and split the three subnets repeatedly. Experimental results indicate that our approach can lead to performance improvement in detecting adaptive steganography.http://dx.doi.org/10.1155/2017/2314860 |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Donghui Hu Qiang Shen Shengnan Zhou Xueliang Liu Yuqi Fan Lina Wang |
spellingShingle |
Donghui Hu Qiang Shen Shengnan Zhou Xueliang Liu Yuqi Fan Lina Wang Adaptive Steganalysis Based on Selection Region and Combined Convolutional Neural Networks Security and Communication Networks |
author_facet |
Donghui Hu Qiang Shen Shengnan Zhou Xueliang Liu Yuqi Fan Lina Wang |
author_sort |
Donghui Hu |
title |
Adaptive Steganalysis Based on Selection Region and Combined Convolutional Neural Networks |
title_short |
Adaptive Steganalysis Based on Selection Region and Combined Convolutional Neural Networks |
title_full |
Adaptive Steganalysis Based on Selection Region and Combined Convolutional Neural Networks |
title_fullStr |
Adaptive Steganalysis Based on Selection Region and Combined Convolutional Neural Networks |
title_full_unstemmed |
Adaptive Steganalysis Based on Selection Region and Combined Convolutional Neural Networks |
title_sort |
adaptive steganalysis based on selection region and combined convolutional neural networks |
publisher |
Hindawi-Wiley |
series |
Security and Communication Networks |
issn |
1939-0114 1939-0122 |
publishDate |
2017-01-01 |
description |
Digital image steganalysis is the art of detecting the presence of information hiding in carrier images. When detecting recently developed adaptive image steganography methods, state-of-art steganalysis methods cannot achieve satisfactory detection accuracy, because the adaptive steganography methods can adaptively embed information into regions with rich textures via the guidance of distortion function and thus make the effective steganalysis features hard to be extracted. Inspired by the promising success which convolutional neural network (CNN) has achieved in the fields of digital image analysis, increasing researchers are devoted to designing CNN based steganalysis methods. But as for detecting adaptive steganography methods, the results achieved by CNN based methods are still far from expected. In this paper, we propose a hybrid approach by designing a region selection method and a new CNN framework. In order to make the CNN focus on the regions with complex textures, we design a region selection method by finding a region with the maximal sum of the embedding probabilities. To evolve more diverse and effective steganalysis features, we design a new CNN framework consisting of three separate subnets with independent structure and configuration parameters and then merge and split the three subnets repeatedly. Experimental results indicate that our approach can lead to performance improvement in detecting adaptive steganography. |
url |
http://dx.doi.org/10.1155/2017/2314860 |
work_keys_str_mv |
AT donghuihu adaptivesteganalysisbasedonselectionregionandcombinedconvolutionalneuralnetworks AT qiangshen adaptivesteganalysisbasedonselectionregionandcombinedconvolutionalneuralnetworks AT shengnanzhou adaptivesteganalysisbasedonselectionregionandcombinedconvolutionalneuralnetworks AT xueliangliu adaptivesteganalysisbasedonselectionregionandcombinedconvolutionalneuralnetworks AT yuqifan adaptivesteganalysisbasedonselectionregionandcombinedconvolutionalneuralnetworks AT linawang adaptivesteganalysisbasedonselectionregionandcombinedconvolutionalneuralnetworks |
_version_ |
1725169386300375040 |