QoE model of scalable MDC stereoscopic video over IP networks

Multiple description coding and path diversity is known to improve the perceptual quality of the received video. This paper considers a machine learning based technique to predict the perceived video quality, in terms of Mean Opinion Score for MDC Stereoscopic video. The perceived experience is expressed as a function of the three-dimensional video representation (color-plus-depth and left-right), the path diversity characterized by asymmetric packet losses between two available paths and the interpolation using a single description delivered over a reliable link as opposed to using multiple unreliable delivered descriptions.