A packet-layer video quality assessment (VQA) model is a lightweight model that predicts the video quality\r\nimpacted by network conditions and coding configuration for application scenarios such as video system planning\r\nand in-service video quality monitoring. It is under standardization in ITU-T Study Group (SG) 12. In this article, we\r\nfirst differentiate the requirements for VQA model from the two application scenarios, and state the argument that\r\nthe dataset for evaluating the quality monitoring model should be more challenging than that for system planning\r\nmodel. Correspondingly, different criteria and approaches are used for constructing the test datasets, for system\r\nplanning (dataset-1) and for video quality monitoring (dataset-2), respectively. Further, we propose a novel video\r\nquality monitoring model by estimating the spatiotemporal complexity of video content. The model takes into\r\naccount the interactions among content features, the error concealment effectiveness, and error propagation\r\neffects. Experiment results demonstrate that the proposed model achieves robust performance improvement\r\ncompared with the existing peer VQA metrics on both dataset-1 and dataset-2. It is noted that on the more\r\nchallenging dataset-2 for video quality monitoring, we obtain a large increase in Pearson correlation from 0.75 to\r\n0.92 and a decrease in the modified RMSE from 0.41 to 0.19.
Loading....