Abstract:
The art of image steganography involves concealing data—text, images, or even
videos—within a cover image. The secret information is concealed such that it cannot
be seen by human sight. Recently, there has been more focus on deep learning
technology, which has proven to be an effective tool in many fields, including image
steganography. The primary objective of this study is to investigate and discuss the
various deep learning techniques that are used in the field of image steganography.
Traditional approaches, Convolutional Neural Network-based methods, and General
Adversarial Network-based methods are the three basic categories into which deep
learning techniques used for image steganography can be separated. This paper
includes a detailed overview of the approach as well as a list of the datasets used,
experimental setups taken into account, and regularly employed evaluation criteria.
