Abstract :
Image steganography is an act of concealing secret information using the image as a cover medium. It is said to be reversible when the same level of importance placed on the retrieval of the secret information is also placed on the recovery of the cover image. The process of hiding information in a cover is called steganography while retrieving the information that was hidden using steganography is called steganalysis. Image steganography is faced with challenges in payload capacity, security, and robustness. Attempts have been made to bring a good solution to this problem but end with a trade-oof in the payload capacity and the security. This paper attempts to solve this problem by proposing a Hybrid Deep Learning Model, which comprises DNN, CycleGAN, and CNN deep learning tools. The study's outcome yielded a good payload capacity and a good security measure, which was evaluated using PSNR and SSIM.
Keyword :
SSIM, CycleGAN, Payload, Security, StegoExpose.