Intelligent Video QoE Prediction Model for Errorprone Networks

Abstract

Objective: Quantifying the quality of video transmitted over diversified networks is becoming critical in the frontiers of digital video Communication. The exponential growth of video-driven applications in wireless domain have created plethora of research issues to be explored in real time. Estimation of video quality in wireless environment requires the conception of better framework and methodologies to improve users’ Quality of Experience (QoE).This paper depicts a novel QoE prediction model that uses machine learning algorithms in predicting video quality over an error prone environment with better prediction accuracy. Methods: The work deploys Pseudo Subjective Quality Assessment (PSQA) method that involves a hybrid technique in assessing the multimedia quality using WEKA machine learning workbench. Findings: The proposed model provides a comparative study of well-known Artificial Intelligence (AI) techniques in predicting the perceived quality of multimedia and the efficiency is analyzed using performance indicators such as Root Mean Squared Error, Correlation Coefficient and Mean Absolute Error. The results of the proposed method underline the effective and inherent advantage in using machine learning methods for video quality prediction. Application: The evaluation results exemplify the importance of applying machine learning paradigms in broad areas of visual quality assessment. The output of various error measures and error variation analysis of the model shows the superiority of Multilayer perceptron based AI technique over other methods.