Subjective video quality estimation to determine optimal spatio-temporal resolution

Video quality improvement plays an important quality of experience (QoE) role in video application. In recent video distributions on the Internet, bitrate is changed automatically by using HTTP adaptive streaming (HAS). Bitrate change often depends on Internet bandwidth, display size, and memory buffer. In each bitrate, resolutions and frame rates can be selected freely. However, it is difficult to compare subjective video quality between different resolutions and frame rates. Therefore, we cannot determine the optimal resolution and frame rate that maximize subjective video quality. In this paper, we propose a method of estimating subjective video quality with various spatial and temporal resolutions. Our method does not need to encode videos for the most part, except for a few to estimate the relationship between bitrate and the quantization parameter (QP).