A multi-task learning CNN for image steganalysis

Xiangyu Yu, Huabin Tan, Hui Liang, Chang Tsun Li, Guangjun Liao

Research output: Book chapter/Published conference paperConference paper

Abstract

Convolutional neural network (CNN) based image steganalysis are increasingly popular because of their superiority in accuracy. The most straightforward way to employ CNN for image steganalysis is to learn a CNN-based classifier to distinguish whether secret messages have been embedded into an image. However, it is difficult to learn such a classifier because of the weak stego signals and the limited useful information. To address this issue, in this paper, a multi-task learning CNN is proposed. In addition to the typical use of CNN, learning a CNN-based classifier for the whole image, our multi-task CNN is learned with an auxiliary task of the pixel binary classification, estimating whether each pixel in an image has been modified due to steganography. To the best of our knowledge, we are the first to employ CNN to perform the pixel-level classification of such type. Experimental results have justified the effectiveness and efficiency of the proposed multi-task learning CNN.

Original languageEnglish
Title of host publication10th IEEE International Workshop on Information Forensics and Security, WIFS 2018
PublisherIEEE, Institute of Electrical and Electronics Engineers
ISBN (Electronic)9781538665367
DOIs
Publication statusPublished - 30 Jan 2019
Event10th IEEE International Workshop on Information Forensics and Security, WIFS 2018 - Hong Kong, Hong Kong
Duration: 10 Dec 201813 Dec 2018

Publication series

Name10th IEEE International Workshop on Information Forensics and Security, WIFS 2018

Conference

Conference10th IEEE International Workshop on Information Forensics and Security, WIFS 2018
CountryHong Kong
CityHong Kong
Period10/12/1813/12/18

Fingerprint

neural network
Neural networks
learning
Classifiers
Pixels
Steganography
efficiency
Classifier

Cite this

Yu, X., Tan, H., Liang, H., Li, C. T., & Liao, G. (2019). A multi-task learning CNN for image steganalysis. In 10th IEEE International Workshop on Information Forensics and Security, WIFS 2018 [8630766] (10th IEEE International Workshop on Information Forensics and Security, WIFS 2018). IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/WIFS.2018.8630766
Yu, Xiangyu ; Tan, Huabin ; Liang, Hui ; Li, Chang Tsun ; Liao, Guangjun. / A multi-task learning CNN for image steganalysis. 10th IEEE International Workshop on Information Forensics and Security, WIFS 2018. IEEE, Institute of Electrical and Electronics Engineers, 2019. (10th IEEE International Workshop on Information Forensics and Security, WIFS 2018).
@inproceedings{36a95c76d85b49ce9e6fba6edd3ae4e9,
title = "A multi-task learning CNN for image steganalysis",
abstract = "Convolutional neural network (CNN) based image steganalysis are increasingly popular because of their superiority in accuracy. The most straightforward way to employ CNN for image steganalysis is to learn a CNN-based classifier to distinguish whether secret messages have been embedded into an image. However, it is difficult to learn such a classifier because of the weak stego signals and the limited useful information. To address this issue, in this paper, a multi-task learning CNN is proposed. In addition to the typical use of CNN, learning a CNN-based classifier for the whole image, our multi-task CNN is learned with an auxiliary task of the pixel binary classification, estimating whether each pixel in an image has been modified due to steganography. To the best of our knowledge, we are the first to employ CNN to perform the pixel-level classification of such type. Experimental results have justified the effectiveness and efficiency of the proposed multi-task learning CNN.",
author = "Xiangyu Yu and Huabin Tan and Hui Liang and Li, {Chang Tsun} and Guangjun Liao",
year = "2019",
month = "1",
day = "30",
doi = "10.1109/WIFS.2018.8630766",
language = "English",
series = "10th IEEE International Workshop on Information Forensics and Security, WIFS 2018",
publisher = "IEEE, Institute of Electrical and Electronics Engineers",
booktitle = "10th IEEE International Workshop on Information Forensics and Security, WIFS 2018",
address = "United States",

}

Yu, X, Tan, H, Liang, H, Li, CT & Liao, G 2019, A multi-task learning CNN for image steganalysis. in 10th IEEE International Workshop on Information Forensics and Security, WIFS 2018., 8630766, 10th IEEE International Workshop on Information Forensics and Security, WIFS 2018, IEEE, Institute of Electrical and Electronics Engineers, 10th IEEE International Workshop on Information Forensics and Security, WIFS 2018, Hong Kong, Hong Kong, 10/12/18. https://doi.org/10.1109/WIFS.2018.8630766

A multi-task learning CNN for image steganalysis. / Yu, Xiangyu; Tan, Huabin; Liang, Hui; Li, Chang Tsun; Liao, Guangjun.

10th IEEE International Workshop on Information Forensics and Security, WIFS 2018. IEEE, Institute of Electrical and Electronics Engineers, 2019. 8630766 (10th IEEE International Workshop on Information Forensics and Security, WIFS 2018).

Research output: Book chapter/Published conference paperConference paper

TY - GEN

T1 - A multi-task learning CNN for image steganalysis

AU - Yu, Xiangyu

AU - Tan, Huabin

AU - Liang, Hui

AU - Li, Chang Tsun

AU - Liao, Guangjun

PY - 2019/1/30

Y1 - 2019/1/30

N2 - Convolutional neural network (CNN) based image steganalysis are increasingly popular because of their superiority in accuracy. The most straightforward way to employ CNN for image steganalysis is to learn a CNN-based classifier to distinguish whether secret messages have been embedded into an image. However, it is difficult to learn such a classifier because of the weak stego signals and the limited useful information. To address this issue, in this paper, a multi-task learning CNN is proposed. In addition to the typical use of CNN, learning a CNN-based classifier for the whole image, our multi-task CNN is learned with an auxiliary task of the pixel binary classification, estimating whether each pixel in an image has been modified due to steganography. To the best of our knowledge, we are the first to employ CNN to perform the pixel-level classification of such type. Experimental results have justified the effectiveness and efficiency of the proposed multi-task learning CNN.

AB - Convolutional neural network (CNN) based image steganalysis are increasingly popular because of their superiority in accuracy. The most straightforward way to employ CNN for image steganalysis is to learn a CNN-based classifier to distinguish whether secret messages have been embedded into an image. However, it is difficult to learn such a classifier because of the weak stego signals and the limited useful information. To address this issue, in this paper, a multi-task learning CNN is proposed. In addition to the typical use of CNN, learning a CNN-based classifier for the whole image, our multi-task CNN is learned with an auxiliary task of the pixel binary classification, estimating whether each pixel in an image has been modified due to steganography. To the best of our knowledge, we are the first to employ CNN to perform the pixel-level classification of such type. Experimental results have justified the effectiveness and efficiency of the proposed multi-task learning CNN.

UR - http://www.scopus.com/inward/record.url?scp=85062862690&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85062862690&partnerID=8YFLogxK

U2 - 10.1109/WIFS.2018.8630766

DO - 10.1109/WIFS.2018.8630766

M3 - Conference paper

T3 - 10th IEEE International Workshop on Information Forensics and Security, WIFS 2018

BT - 10th IEEE International Workshop on Information Forensics and Security, WIFS 2018

PB - IEEE, Institute of Electrical and Electronics Engineers

ER -

Yu X, Tan H, Liang H, Li CT, Liao G. A multi-task learning CNN for image steganalysis. In 10th IEEE International Workshop on Information Forensics and Security, WIFS 2018. IEEE, Institute of Electrical and Electronics Engineers. 2019. 8630766. (10th IEEE International Workshop on Information Forensics and Security, WIFS 2018). https://doi.org/10.1109/WIFS.2018.8630766