WGAN-GP Based Generative Coverless Information Hiding

As a prominent research focus in steganographic communication, coverless information hiding aims to fundamentally resist steganography detection algorithms. However, current methods face several challenges, including reliance on large-scale image databases, limited hiding capacity, and poor quality of generated images, which significantly constrain their practical applicability. To address these issues, this paper proposes a generative coverless information hiding method based on WGAN-GP. First, the mapping relationship between secret information and attribute label codes is established. The sender then inputs the secret information along with real images into a generator to produce images with specific styles. These generated images are subsequently combined into a secret image containing the hidden secret and sent to the receiver. The receiver uses a discriminator to analyze the received image and extract the hidden information. Experimental results demonstrate that the proposed method does not require large-scale image databases and excels in terms of image quality, hiding capacity, robustness, and security, offering a more viable solution for practical applications.