PROJECT TITLE :

A Convex Optimization Framework for Video Quality and Resolution Enhancement From Multiple Descriptions

ABSTRACT:

Streaming and compressing methods Of the last decade, technological advancements have led to a migration in multimedia material to the cloud. The same video can be accessed through a variety of distribution channels. To meet the needs of various applications, various compression levels, algorithms, and resolutions are employed. Resolution enhancement techniques will become increasingly important as 4k display technologies become more widely used. Currently available solutions do not account for the unique characteristics of various video encoders, and approaches for reconstructing video from compressed sources do not provide resolution augmentation. Compression levels and resolutions for each description can be varied in this paper's multi-source compressed video improvement approach. Multiple descriptions of the same video can be combined quickly and effectively using a variational formulation built on a recent proximal dual splitting method. LR and HR compressed descriptions of a video sequence can be combined to produce an HR description, and the latter can be used to improve the quality of an HR compressed video using the former. A variety of downsampling methods and compression settings are used to test a variety of video sequences encoded with high efficiency video coding.