Perceptual Quality Assessment of Video Considering Both Frame Rate and Quantization Artifacts

A Study on Quality Metrics for 360 Video Communications

“…However, there has been few proposed quality metrics considering the impacts of content characteristics. In this study, a content-related metric that is proposed in [3] is investigated.…”

“…Hybrid metric using Content Features (HMCF) [3] is calculated based on measured mean opinion score Q 0 (subjective quality metric) of an original video and PSNR (objective quality metric) between the original video and a test video. Specifically, HMCF is calculated by…”

“…While there are a lot of previous studies on traditional videos [1]- [3], the research on 360 videos is still very limited. Most of 360 video related studies focus on investigating some aspects of video quality [4]- [6] in VR environment such as the presence, usability, and cybersickness.…”

“…However, the use and investigation of advanced quality metrics such as structural similarity-related metrics and contentbased quality metrics (e.g. [3], [14], [15]) are still very limited. In this work, we investigate both objective quality and subjective quality of 360 videos.…”

“…Objective quality metrics are then evaluated based on their complexity and correlations with subjective quality. Regarding objective quality metrics, we not only consider the above PSNR-based metrics, but also investigate three other advanced metrics, namely, SSIM [14], MS-SSIM [15], and a hybrid metric using content features (HMCF) [3]. Based on the evaluation results, good objective quality metrics for different processing phases are identified.…”