Multimedia streaming services have been among the top applications on the internet, due to the vast amount of information presented through it. The advent of the semantic web and the services surrounding it includes many methods of accurately describing multimedia content. This paper investigates the use of a deep learning method to improve the quality of multimedia. Deep learning is a class of machine learning methodologies that uses many-layered artificial neural networks to learn internal representations of data and is showing increasing promise in optimizing multimedia services in complex ways. In the proposed intensification methodology, we integrate Deep Learning algorithms into different multimedia services, particularly focusing on content adaptation techniques such as coding, transcoding, and encryption. Then, we elaborate on how the Deep Learning algorithm can be integrated into coding and transcoding services, which are then followed by how it can facilitate intelligent content adaptation in encryption. Finally, we conducted an experimental comparison between traditional machine learning methods and deep learning methodology to automatically induce a physics simulator to demonstrate great future promise in this field. This comprehensive literature review has revealed that image compression remains uncharted territory for content-based methods, while steady progression has been made in other MIR domains.

Multimedia, Deep Learning, Machine Learning, Encryption, Image Compression