Summary
International Conference on Emerging Technologies for Communications
2020
Session Number:E3
Session:
Number:E3-3
GA-based feature selection for QoE estimation using EEG during video viewing
Kasumi Kitao, Daichi Kominami, Masayuki Murata,
pp.-
Publication Date:2020/12/2
Online ISSN:2188-5079
PDF download
Summary:
In recent years, Quality of Experiment (QoE) becomes an important factor for video viewing users and QoE-based video delivery control methods have been getting a lot of interests. In order to use the user's QoE for video delivery control, it is necessary to be able to measure the QoE suitable for the individual in real time. In this paper, we present a support vector machine based estimation method for the QoE of video viewing user using the user's EEG. We extracted over 400 features from the EEG measurements, but we show that the number of features does not need to be so large in the estimation. We also show that the feature selection based on the genetic algorithm(GA) can improve the accuracy of QoE estimation by an average of 6% compared to the random selection.