LIVE-ETRI (ETRI-LIVE Space-Time Subsampled Video Quality (STSVQ) Database)

视频部署参数空间不断增加，为全球流媒体和社交媒体观众提供更逼真和身临其境的体验。然而，空间分辨率或帧速率等视频参数的增加不可避免地与更大的数据量相关联。通过有限带宽网络以感知最佳的方式传输越来越多的视频是影响数十亿观众的当前挑战。视频服务提供商最近采用的一种做法是结合视频压缩的时空分辨率调整。因此，了解不同级别的时空二次采样和压缩如何影响视频的感知质量非常重要。为了朝着这个方向取得进展，我们构建了一个名为 ETRI-LIVE 时空二次采样视频质量 (ETRI-LIVE-STSVQ) 数据库的大型新资源，其中包含通过应用不同级别的组合时空二次采样和视频生成的 437 个视频压缩 15 种不同的视频内容。我们还对新数据集进行了大规模的人体研究，收集了大约 15,000 个视频质量的主观判断。 ETRI-LIVE STSVQ 数据库正在公开和免费提供，希望改进未来对视频质量建模和感知视频编码等主题的研究和开发。

The parameter space for video deployment continues to expand, delivering more realistic and immersive experiences to global streaming and social media audiences. However, increases in video parameters such as spatial resolution or frame rate are inevitably tied to larger data volumes. Transmitting increasingly large volumes of video over bandwidth-limited networks in a perceptually optimal manner is a current challenge affecting billions of viewers. A practice recently adopted by video service providers is to combine spatial-temporal resolution adjustment with video compression. Therefore, it is important to understand how different levels of spatial-temporal subsampling and compression affect the perceptual quality of videos. To advance research in this direction, we constructed a large new resource called the ETRI-LIVE Spatial-Temporal Subsampling Video Quality (ETRI-LIVE-STSVQ) database, which contains 437 compressed videos generated by applying different combinations of spatial-temporal subsampling levels across 15 distinct video contents. We also carried out a large-scale human subject study on this new dataset, collecting approximately 15,000 subjective judgments of video quality. The ETRI-LIVE STSVQ database is publicly and freely available, with the aim of improving future research and development on topics such as video quality modeling and perceptual video coding.

搜集汇总

背景与挑战

背景概述

该数据集旨在探究时空二次采样和压缩对视频感知质量的影响，包含437个基于15种不同内容生成的视频，并收集了约15,000个人类主观质量评分。它作为公开资源，旨在推动视频质量建模和感知编码等领域的研究进展。