A transmission-aware video quality metric for lossy wireless network

In this paper, we propose a Transmission-aware Quality Metric (TQM) for real-time evaluation of video quality in lossy wireless network. There are two major contributions in our study. Firstly, in order to guarantee the real-time requirements, we considers the impacts of Packet Loss Rate (PLR) and Maximum Transmission Unit (MTU) which can be extracted directly from bit-stream, and build TQM by using the relationship of subjective score and these two transmission parameters. Then, the content-aware feature which is achieved from a linear prediction of developed Just Noticeable Distortion (JND), is included in TQM to improve the prediction accuracy by introduce the discrimination of different videos. The verification results show that the predicted video quality measured by Mean Opinion Score (MOS) has a high Pearson correlation for unknown videos under different packet loss schemes (higher than 0.75 in distributed loss and 0.71 in burst loss).