ABSTRACT The advent of the modern technological era has allowed us to reach a stage where people can share their information via different platforms quite easily. These platforms allow users to express themselves through text, photographs, videos, and audio, among other different representational media. The amount of photographic data is higher in comparison with other forms of data. So, the security of these images is a big concern for the researchers. Deep learning (DL)‐based approaches have gained popularity for a variety of multimedia analysis applications, including segmentation, detection, and classification. This article presents a state‐of‐the‐art summarization of DL‐based multimedia security techniques in which various encryption techniques, covert operations, current challenges, the scope of improvement, and new directions are highlighted. This paper mentions a comprehensive review of different DL watermarking and hiding techniques, along with a comparison of the contributions of the literature. This survey, in our opinion, can open the door to further investigating the essential topic of information concealment in DL environments.
Awasthi et al. (Wed,) studied this question.