Abstract:As technology advances, the steganalysis methods based on Convolutional Neural Network (CNN)can more precisely detect steganography behavior. However, the existing CNN-based steganalysis methods have the problem of insufficient robustness in adversarial environments. Specifically, the steganography methods take advantage of adversarial examples' capacity to deceive deep learning models, adding small perturbations to the stego images to generate adversarial stego images that can successfully evade the detection of CNN-based steganalysis. In this work, we design a robust steganalysis based on the Generative Adversarial Network (GAN), namely RS-GAN. The proposed method reconstructs the suspicious images after learning the distribution of clean images to identify adversarial stego images. Specifically, RS-GAN first disentangles suspicious images to obtain the texture and structure vectors. Afterward, reconstructed images are synthesized based on the above vectors, which have similar distributions to the clean images. In addition, we design an anomaly score calculation module to obtain the reconstruction loss and the discriminator loss and then determine the anomaly scores. Finally, the RS-GAN compares the anomaly scores with the threshold to eliminate the anomaly examples. According to the experimental comparison and analysis, RS-GAN can effectively eliminate the adversarial stego images and increase the robustness of steganalysis in adversarial environments. In the LSUN dataset, when the Fast Gradient Method is employed as the attack with a perturbation value of 0.0013, the accuracy of YeNet decreased to 47.867%, while RS-GAN maintains an accuracy of 72.100%.

ARES: on Adversarial Robustness Enhancement for Image Steganographic Cost Learning

A Dct-Based Image Steganographic Method Resisting Statistical Attacks

Generative adversarial networks-based image steganography with multiscale features integration

Channel Attention Image Steganography With Generative Adversarial Networks

Improving the robustness of steganalysis in the adversarial environment with Generative Adversarial Network

Spatial Image Steganography Based on Generative Adversarial Network

A Novel Grayscale Image Steganography via Generative Adversarial Network

On Security Enhancement of Steganography via Generative Adversarial Image

Robustness enhancement against adversarial steganography via steganalyzer outputs?

A stable GAN for image steganography with multi-order feature fusion

A High-Capacity Steganography Algorithm Based on Adaptive Frequency Channel Attention Networks

A Novel Approach to Image Steganography Using Generative Adversarial Networks

LDStega: Practical and Robust Generative Image Steganography Based on Latent Diffusion Models

Sstegan: Self-Learning Steganography Based On Generative Adversarial Networks

A Novel Gradient-guided Post-processing Method for Adaptive Image Steganography

An Effective Imbalanced JPEG Steganalysis Scheme Based on Adaptive Cost-Sensitive Feature Learning

GAN-Based Steganography with the Concatenation of Multiple Feature Maps.

Highly Accurate End-to-end Image Steganalysis Based on Auxiliary Information and Attention Mechanism

Deeply‐Recursive Attention Network for Video Steganography

Improving image steganography security via ensemble steganalysis and adversarial perturbation minimization

Adversarial feature hybrid framework for steganography with shifted window local loss