Learning Symmetric and Asymmetric Steganography via Adversarial Training

TL;DR

This paper introduces a novel adversarial training-based steganographic scheme that improves message concealment and detection resistance, achieving higher invisibility and undetectability compared to previous methods, with successful real-world simulation results.

Contribution

The paper presents the first key-dependent adversarial training schemes for symmetric and asymmetric steganography, enhancing concealment and detection resistance.

Findings

20% improvement in invisibility over previous deep-learning methods

25% better undetectability than classic algorithms

98% message decoding accuracy in real-world simulation

Abstract

Steganography refers to the art of concealing secret messages within multiple media carriers so that an eavesdropper is unable to detect the presence and content of the hidden messages. In this paper, we firstly propose a novel key-dependent steganographic scheme that achieves steganographic objectives with adversarial training. Symmetric (secret-key) and Asymmetric (public-key) steganographic scheme are separately proposed and each scheme is successfully designed and implemented. We show that these encodings produced by our scheme improve the invisibility by 20% than previous deep-leanring-based work, and further that perform competitively remarkable undetectability 25% better than classic steganographic algorithms. Finally, we simulated our scheme in a real situation where the decoder achieved an accuracy of more than 98% of the original message.