Region Priority Based Adaptive 360-Degree Video Streaming Using DASH

Note: We don't have the ability to review paper

PubDate: September 2018

Teams: Chongqing University of Posts and Telecommunications

Writers: Fuxing Yang; Jiangtao Luo; Junchao Yang; Zheng Xu

PDF: Region Priority Based Adaptive 360-Degree Video Streaming Using DASH


Abstract

With the continuous improvement of Virtual Reality (VR) hardware and software facilities and development of VR streaming platform, the VR industry ushered in a period of rapid development. VR makes use of 360-degree panoramic or omnidirectional video with high resolution and high frame rate in order to create the immersive experience to the user. However those characteristics cause the big volume of 360-degree video and bandwidth intensive during transmission. Due to the limitation of the human eye vision, the user can only watch the part of the 360-degree video in a head-mounted display (HMD) at one time. Hence, streaming the VR video by the traditional video transmission method causes bandwidth waste. In view of this, this paper proposed a 360-degree video adaptive transmission method based on user viewport. Firstly, a 360-degree video region prioritization scheme is proposed based on user vision characteristics and head motion features. Different priority regions transmit 360-degree video content with different quality levels as the format of tiles. Then, based on the bandwidth estimation, buffer status and user viewport prediction, a 360-degree video adaptive transmission decision strategy is given. According to the predicted available bandwidth and viewport, the quality combination of the 360-degree video tiles to be transmitted in the near future is determined to implement adaptive transmission. The test results of the 360-degree video transmission experiment system based on the DASH standard show that the proposed 360-degree video adaptive transmission strategy can effectively reduce the bandwidth consumption on the basis of guaranteeing the user Quality of Experience(QoE).

更多阅读推荐......

Paper