A perceptually-based objective measure for speech coders using abductive network

Objective measure techniques that provide reliable performance evaluation of speech coders is indispensable to designers, to end users when selecting telecommunication services and to providers in the designing stage of the telecommunications facilities. We introduce our objective measure technique that: (1) emulates several known features of perceptual processing of speech sounds by the human ear to map the speech power spectrum into an auditory power spectrum that is used to calculate the bark spectral distance per band (BSDB) between the input and the output coded speech signals; and (2) uses the abductive networks to estimate the speech quality from the BSDB. The results indicate that our proposed technique is reliable and robust in evaluating the coded speech quality and it is highly correlated to human responses across a wide range of quality levels and for a wide range of speech processing.