A distributed delivery model for 3D-video streams

3D-video systems allow a user to perceive depth in the viewed scene and to interactively display the scene from multiple viewpoints. This paper presents an architecture for the scalable delivery of 3D-video streams over IP-networks, where we contribute in two aspects. First, we model 3D-video as a multi-stream application, where each stream (or layer) carries a single coded video signal or coded scene-description data. Second, we propose an end-to-end delivery model for 3D-video applications, which leverages a distributed system architecture to reduce the bandwidth and processing cost at the server and the end-hosts.