Advanced intelligent data hiding using video stego and hybrid convolutional neural networks

The practice of steganography involves the concealment of confidential data inside other, seemingly innocuous files of the same or other sorts. The objective of this study is to create a stego technique that, when applied to a video clip, will successfully conceal a message inside its graphics. A model is developed for video steganography by developing a model to conceal video inside another video using hybrid convolutional neural networks (HyCNN). The second objective is to expand the size of the space that can be used for hiding, which has been accomplished via the use of CNN. The suggested model was trained using HyCNN on arbitrary pictures drawn from the ImageNet database. The findings also show that the system is able to produce excellent results in visibility and attacks, where the suggested approach is able to effectively mislead both the observer and the steganalysis software.