Secret-to-Image Reversible Transformation for Generative Steganography

• Published in: IEEE transactions on dependable and secure computing Vol. 20; no. 5; pp. 1 - 17
• Main Authors: Zhou, Zhili, Su, Yuecheng, Li, Jin, Yu, Keping, Jonathan Wu, Q. M., Fu, Zhangjie, Shi, Yunqing
• Format: Journal Article
• Language: English
• Published: Washington IEEE 01.09.2023; IEEE Computer Society
• Subjects: coverless steganography; Data mining; digital forensics; Distortion; Encoding; Feature extraction; Generative adversarial networks; generative steganography; information hiding; Information retrieval; Messages; Normal distribution; Steganography; Transformations (mathematics); Transforms
• ISSN: 1545-5971; 1941-0018
• DOI: 10.1109/TDSC.2022.3217661

Recently, generative steganography that transforms secret information to a generated image has been a promising technique to resist steganalysis detection. However, due to the inefficiency and irreversibility of the secret-to-image transformation, it is hard to find a good trade-off between the information hiding capacity and extraction accuracy. To address this issue, we propose a secret-to-image reversible transformation (S2IRT) scheme for generative steganography. The proposed S2IRT scheme is based on a generative model, i.e., Glow model, which enables a bijective-mapping between latent space with multivariate Gaussian distribution and image space with a complex distribution. In the process of S2I transformation, guided by a given secret message, we construct a latent vector and then map it to a generated image by the Glow model, so that the secret message is finally transformed to the generated image. Owing to good efficiency and reversibility of S2IRT scheme, the proposed steganographic approach achieves both high hiding capacity and accurate extraction of secret message from generated image. Furthermore, a separate encoding-based S2IRT (SE-S2IRT) scheme is also proposed to improve the robustness to common image attacks. The experiments demonstrate the proposed steganographic approaches can achieve high hiding capacity (up to 4 bpp ) and accurate information extraction (almost 100% accuracy rate) simultaneously, while maintaining desirable anti-detectability and imperceptibility.