Optimized spatial and temporal resolution based on subjective quality estimation without encoding

Abstract

In this paper, we propose a method of estimating subjective video quality with various spatial and temporal resolutions without encoding. Under a given bitrate constraint, the combination of resolution and frame rate that provides best subjective video quality depends on the video content. To maximize subjective video quality, several studies have proposed models that can estimate subjective quality with various resolutions and frame rates. However, to determine the optimal resolution and frame rate that maximize subjective video quality, it is necessary to estimate subjective video quality at each combination of resolution/frame rate/bitrate. This takes considerable time with previously reported methods because they require an encoding process for decoding videos or obtaining pre analysis. To address this issue, we developed a method that does not require an encoding process to estimate subjective video quality.