Inferring Video QoE in Real Time

Inferring the subjective perception of a video stream in real time continues to be a stiff problem. This article presents MintMOS: a lightweight, no-reference, loadable kernel module to infer the QoE of a video stream in transit and offer suggestions to improve it. MintMOS revolves around one-time offline construction of a k-dimensional space, which we call the QoE space. A QoE space is a known characterization of subjective perception for any k parameters (network dependent/independent) that affect it. We create N partitions of the QoE space by generating N video samples for various values of the k parameters and conducting subjective surveys using them. Every partition then has an expected QoE associated with it. Instantaneous parameters of a real-time video stream are compared to the precomputed QoE space to both infer and offer suggestions to improve QoE. Inferring QoE is a lightweight algorithm that runs in linear time. We implemented MintMOS by creating an actual QoE space using three parameters and 27 partitions by conducting surveys with 77 human subjects. In a second set of surveys using 13 video clips, MintMOS\u27s predictions were compared to 49 human responses. Results show that our MOS predictions are in close agreement with subjective perceptions