Delay-Distortion Optimization for Content-Adaptive Video Streaming

We propose a new pre-roll delay-distortion optimization (DDO) framework that allows determination of the minimum pre-roll delay and distortion while ensuring continuous playback for on-demand content-adaptive video streaming over limited bitrate networks. The input video is first divided into temporal segments, which are assigned a relevance weight and a maximum distortion level, called relevance-distortion policy, which may be specified by the user. The system then encodes the input video according to the specified relevance-distortion policy, whereby the optimal spatial and temporal resolutions and quantization parameters, also called encoding parameters, are selected for each temporal segment. The optimal encoding parameters are computed using a novel, multi-objective optimization formulation, where a relevance weighted distortion measure and pre-roll delay are jointly minimized under maximum allowable buffer size, continuous playback, and maximum allowable distortion constraints. The performance of the system has been demonstrated for on-demand streaming of soccer videos with substantial improvement in the weighted distortion without any increase in pre-roll delay over a very low-bitrate network using AVC/H.264 encoding

I. Introduction

Pre-Roll delay is a vital parameter in video streaming since it provides some level of protection against network throughput variations, as well as allowing flexible rate allocation in video coding. If it is chosen too small, pauses in video playback due to network throughput variations and/or unacceptable video quality due to strict rate control in video coding would result. An unnecessarily large pre-roll delay, which in the limit leads to the download-and-play solution, requires a very long initial wait, thus eliminating the benefit of streaming, and is usually found objectionable by users. Therefore, video streaming applications should strike the right balance between pre-roll delay and video distortion. This issue becomes even more significant in content-adaptive video streaming over low-bitrate networks, where different bitrates (sometimes larger than the network throughput) shall be allocated to different temporal video segments (shots) according to their importance.