Visual saliency driven error protection for 3D video

Viewers tend to focus into specific regions of interest in an image. Therefore visual attention is one of the major aspects to understand the overall Quality of Experience (QoE) and user perception. Visual attention models have emerged in the recent past to predict user attention in images, videos and 3D video. However, the usage of these models in quality assessment and quality improvement has not been thoroughly investigated to date. This paper investigates 3D visual attention model driven quality assessment and improvement methods for 3D video services. Moreover, a visual saliency driven error protection mechanism is proposed and evaluated in this paper. Both objective and subjective results show that the proposed method has significant potential to provide improved 3D QoE for end users.