Steganographic Generative Adversarial Networks

被引：90

作者：

Volkhonskiy, Denis ^{[1
]}

Nazarov, Ivan ^{[1
]}

Burnaev, Evgeny ^{[1
]}

机构：

[1] Skolkovo Inst Sci & Technol, Nobel St 3, Moscow, Moskovskaya Obl, Russia

来源：

TWELFTH INTERNATIONAL CONFERENCE ON MACHINE VISION (ICMV 2019) | 2020年 / 11433卷

关键词：

generative adversarial networks; steganography; security;

D O I：

10.1117/12.2559429

中图分类号：

O43 [光学];

学科分类号：

070207 ; 0803 ;

摘要：

Steganography is collection of methods to hide secret information ("payload") within non-secret information "container"). Its counterpart, Steganalysis, is the practice of determining if a message contains a hidden payload, and recovering it if possible. Presence of hidden payloads is typically detected by a binary classifier. In the present study, we propose a new model for generating image-like containers based on Deep Convolutional Generative Adversarial Networks (DCGAN). This approach allows to generate more setganalysis-secure message embedding using standard steganography algorithms. Experiment results demonstrate that the new model successfully deceives the steganography analyzer, and for this reason, can be used in steganographic applications.

引用

页数：15