Optimizing 360° Video Streaming to Head-Mounted Virtual Reality
PubDate: October 2018
Teams: National Tsing Hua University
Writers: Ching-Ling Fan;Hsin-Chu; Cheng-Hsin Hsu
PDF: Optimizing 360° Video Streaming to Head-Mounted Virtual Reality
Abstract
Streaming 360 ° videos to Head-Mounted Displays (HMDs) is very challenging due to large video sizes, stringent real-time requirements, and complex human visual systems. In my PhD study, we focus on three core problems to optimize 360 ° video streaming to HMDs. First, we adopt sensor and content data to design a fixation prediction algorithm. Second, we develop a user study to construct a 360° Quality-of-Experience (QoE) model. Last, we design, implement, and evaluate an optimal joint tile selection and bitrate allocation algorithm that leverages both the fixation prediction algorithm and QoE model. Initial results are reported in this paper.