Paper
5 May 2022 Generative image steganography based on VAE-GAN
Pan Yang, Minqing Zhang, Yu Ge, Yingnan Zhang
Author Affiliations +
Proceedings Volume 12245, International Conference on Cryptography, Network Security, and Communication Technology (CNSCT 2022); 122450I (2022) https://doi.org/10.1117/12.2635940
Event: International Conference on Cryptography, Network Security, and Communication Technology (CNSCT 2022), 2022, Sanya, China
Abstract
Aiming at the problems of distortion, low quality, and multiple transmissions of generative stego-image based on generative adversarial networks, a new generative image steganography method combing variational autoencoders (VAE) and generative adversarial networks (GAN) is proposed. This method combines the advantages of VAE and GAN. First, the binary secret information is grouped, and flag bits are added, which is converted into noise according to the mapping relationship. The noise is input into the VAE-GAN generator to obtain a high-quality stego-image group and randomly combine a new image. When extracting information, the stego-image is input to the extractor to recover the noise and converted into a binary message according to the mapping relationship. Finally, the secret information is recovered according to the flag bit. Experimental results show that the method not only improves the quality of stego-image but also has a high steganography capacity.
© (2022) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Pan Yang, Minqing Zhang, Yu Ge, and Yingnan Zhang "Generative image steganography based on VAE-GAN", Proc. SPIE 12245, International Conference on Cryptography, Network Security, and Communication Technology (CNSCT 2022), 122450I (5 May 2022); https://doi.org/10.1117/12.2635940
Advertisement
Advertisement
RIGHTS & PERMISSIONS
Get copyright permission  Get copyright permission on Copyright Marketplace
KEYWORDS
Steganography

Image quality

Data hiding

Computer programming

Image transmission

Binary data

Distributed interactive simulations

Back to Top