Distortion-Aware Virtual Environments Transmission Scheme Over A Bandwidth-Limited Network

Virtual environments (VEs) have been widely used in multimedia applications, with trends toward larger data size. 3D streaming has recently been proposed to enable users to view the VEs in real time. However, how to efficiently download the VEs over a bandwidth-limited network becomes a bottleneck. With the aim of minimizing the scene distortion and achieving a more efficient use of bandwidth, it's necessary to download the visible 3D contents at an appropriate resolution. In this paper, we propose a distortion-aware VEs transmission scheme. The main contributions of the proposed scheme include: (1) Improving the traditional interest management algorithm, so that our method can fit the user's visual and behavioral characteristics better. (2) Proposing a scene distortion estimation algorithm, by which we can effectively calculate and evaluate the scene distortion. (3) Based on (1) and (2), we propose a scene update strategy, which combines the user's visual characteristics, scene distortion and bandwidth estimation. The experimental results show that the proposed scheme can lower distortion especially when network condition is poor.