Objective measures for speech quality assessment in wireless communications

Automatic assessment of speech quality is one of the most important problems in mobile communication. The primary goal of this article is to find an objective measure that accurately predicts user acceptance of speech signals recorded under varying noise conditions and channel impairments as well as signals subjected to various noise suppression/signal enhancement techniques. We report some preliminary results on correlation studies between several output-based/input-output-based objective measures and subjective scores. We propose to use these results to design a non-linear model for predicting the subjective quality using several simple objective measures