An improved QoE estimation method based on QoS and affective computing

Abstract

With the massive uses of the video over the world in the last decade, the user perception, commonly called Quality of Experience (QoE) metric; has become the one of the most important topics for the Network services Providers (NsP) and Content service Providers (CsP). In this paper, we present a new QoE estimation method on the client side using Machine Learning methods (ML) based on subjective assessment in a controlled-laboratory environment. The major novel contribution of this study is the combination of Quality of Service (QoS) parameters and Affective Computing (facial expression) to estimate the Mean Opinion Score (MOS) for HTTP YouTube content. An evaluation using a collected subjective dataset indicates that combining QoS and Affective computing provides better prediction performance.