ARES: On Adversarial Robustness Enhancement for Image Steganographic Cost Learning
- 12 January 2024
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 26 (15209210) , 6542-6553
Abstract
Taking the steganalytic discriminators as the adversaries, the existing Generative Adversarial Networks (GAN)-based steganographic approaches learn the implicit cost functions to measure the embedding distortion for steganography. However, the steganalytic discriminators in these approaches are trained by the stego-samples with insufficient diversity, and their network structures offer very limited representational capacity. As a result, these steganalytic discriminators will not exhibit robustness to various steganographic patterns, which causes learning suboptimal cost functions, thus compromising the anti-steganalysis capability. To address this issue, we propose a novel GAN-based steganographic approach, in which the Diversified Inverse-Adversarial Training (DIAT) strategy and the Steganalytic Feature Attention (SteFA) structure are designed to train a robust steganalytic discriminator. Specifically, the DIAT strategy provides the steganalytic discriminator with an expanded feature space by generating diversified adversarial stego-samples; the SteFA structure enables the steganalytic discriminator to capture more various steganalytic features by employing the channel-attention mechanism on higher-order statistics. Consequently, the steganalytic discriminator can build a more precise decision boundary to make it more robust, which facilitates learning a superior steganographic cost function. Extensive experiments demonstrate that the proposed steganographic approach achieves promising anti-steganalysis capability over the state-of-the-arts under the same embedding payloads.Keywords
Funding Information
- National Natural Science Foundation of China (62372125, 61972205, U1936218)
- Guangdong Natural Science Funds for Distinguished Young Scholar (2023B1515020041)
- China Scholarship Council (202008320533, 202109040027)
- Guangzhou Research Foundation (202201010194, 202201020139)
This publication has 49 references indexed in Scilit:
- A stable GAN for image steganography with multi-order feature fusionNeural Computing and Applications, 2022
- High frequency patterns play a key role in the generation of adversarial examplesNeurocomputing, 2021
- Research on covert communication channel based on modulation of common compressed speech codecNeural Computing and Applications, 2020
- GAN-Based Steganography with the Concatenation of Multiple Feature Maps2020
- Adaptive spatial steganography based on adversarial examplesMultimedia Tools and Applications, 2019
- A Steganography Algorithm Based on CycleGAN for Covert Communication in the Internet of ThingsIEEE Access, 2019
- Image Steganography Based on Foreground Object Generation by Generative Adversarial Networks in Mobile Edge Computing With Internet of ThingsIEEE Access, 2019
- U-Net: Convolutional Networks for Biomedical Image Segmentation2015
- ”Break Our Steganographic System”: The Ins and Outs of Organizing BOSS2011
- Using High-Dimensional Image Models to Perform Highly Undetectable Steganography2010