VP8 Video : Performance Evaluation, Modeling and Frame Size Prediction
Abstract
Recently released open source video codec VP8 attract researchers from academia and industry. This is mainly due to the better compression efficiency and low traffic variability of VP8 along with achieving higher video quality as compared to other popular video codecs like H.264/AVC. Applications like video streaming and video calling are rapidly increasing.This provides an opportunity to carefully select the video codec and manage network resources. In this paper VP8 and H.264/AVC are compared for compression efficiency, traffic variability and objective video quality for similar encoding parameters, VP8 outperforms H.264 in our experiments.We extend our work regarding modeling and prediction of VP8 encoded video. Seasonal Autoregressive Integrated Moving Average (SARIMA) method is used for modeling the video frame trace and based on the obtained parameters frame size is predicted.We find that the model accurately represents all video frame traces listed in this paper. Encoding methodology for VP8 video is also discussed. Moreover, we compare SARIMA with other models used for predicting video frame trace and show its suitability in terms of fitness of model for our purpose. We observe Akaike information criterion (AIC) value for SARIMA is 3662.26 as compared to ARIMA with 4742.90 and Exponential smoothing with 5430.16.The model for VP8 video prediction in advance can be used for optimization and resource allocation over broadband network more efficiently
Award : 2011 IEEE R10 Best Paper contest (Postgraduate Section) ( 3rd Best Paper Award)
Featured at : http://www.r10sac.org/activities/results.html : 2011
©2011 IEEE. Personal use of this material is permitted. However, permission to reprint/republish this material for advertising or promotional purposes or for creating new collective works for resale or redistribution to servers or lists, or to reuse any copyrighted component of this work in other works must be obtained from the IEEE