Video Quality Assessment

Video Quality Assessment is a computer vision task aimed at modeling human subjective perception based on videos, with the goal of generating a Mean Opinion Score (MOS) where a higher score indicates better perceived quality. This task has significant application value in video coding, transmission, and processing, among other fields. Common benchmark datasets include KoNViD-1k, LIVE-VQC, YouTube-UGC, and LSVQ, and the evaluation metrics are SROCC, PLCC, and RMSE. As a subtask of Time Series, it focuses on the quality analysis of video streams.