Video quality assessment in video streaming services considering user preference for video content

In video streaming service, the user's Quality of Experience (QoE) is not only related to video signal quality received at consumer's devices, the users’ subjectivity must also be considered. In this context, a video quality assessment method that takes into account the user's preference for video content is proposed in this research. In order to perform this task, the users' profiles that include their preferences were stored in a video server. Then, subjective tests of video quality assessment were conducted, in which evaluators had different video content preferences. Results show that the evaluators' QoE is highly correlated with the user's preference for video content type. Based on these experimental results, a function named Preference Factor (PF) is defined and used to adjust the quality index values obtained by an objective video quality metric running in the end user's device. The PF function also depends on video content type and quality index score. Using the PF function, the enhanced Video streaming Quality Metric (e-VsQM) is proposed and the results of its performance evaluation demonstrate that PF improves an objective video quality metric. Furthermore, e-VsQM has low complexity and can be utilized in different video services. Thus, an application scenario is presented, in which the proposed video quality metric is implemented.