As the video streaming industry is transitioning from high-definition to ultra-high-definition resolutions, new techniques must be explored to accommodate for this significant change in video resolution transmission and its impact on wireless networks. This paper explores the application of a machine learning upscaling algorithm to a compression pipeline in which the video is both compressed and downscaled prior to transmission over a wireless channel. Once the video is received by the end user it is then upscaled. The effects of the compression pipeline and wireless packet loss on video quality are considered and measured against traditional interpolation methods which serve as a baseline for measuring upscaling performance.
