Spatial Image Steganography Based on Generative Adversarial Network

TL;DR

This paper introduces a novel spatial image steganography method using a GAN architecture with a U-NET generator, Tanh-simulator, and SCA-aware discriminator, achieving higher security and faster training than existing methods.

Contribution

The paper presents a new GAN-based steganography framework with a U-NET generator, Tanh-simulator, and SCA-aware discriminator, improving security and training efficiency over prior approaches.

Findings

Significantly improves security performance over ASDL-GAN.

Reduces training time to 30% of ASDL-GAN.

Outperforms hand-crafted S-UNIWARD in security.

Abstract

With the recent development of deep learning on steganalysis, embedding secret information into digital images faces great challenges. In this paper, a secure steganography algorithm by using adversarial training is proposed. The architecture contain three component modules: a generator, an embedding simulator and a discriminator. A generator based on U-NET to translate a cover image into an embedding change probability is proposed. To fit the optimal embedding simulator and propagate the gradient, a function called Tanh-simulator is proposed. As for the discriminator, the selection-channel awareness (SCA) is incorporated to resist the SCA based steganalytic methods. Experimental results have shown that the proposed framework can increase the security performance dramatically over the recently reported method ASDL-GAN, while the training time is only 30% of that used by ASDL-GAN. Furthermore, it also performs better than the hand-crafted steganographic algorithm S-UNIWARD.