Study On PESQ Algorithm Based On The Auditory Model

In this paper, the classification and performance of the objective methods were introduced in detail and the PESQ in ITU-T P.862 Recommendation was proposed. And this paper presented the structural drawing of the algorithm and gave the simulation of Perceptually-Based Features for Objective Speech Quality Evaluation. The Intermediate Reference System (IRS) filter was replaced by a band-pass filter to simplify calculation. And weights over speech frames in disturbance integration procedure were reallocated to simulate the mask effect of human hearing. Finally, the simulation-results of PESQ were given. And the correlation-coefficient of PESQ and subjective scores were calculated. The experiment results show that much time complexity can be degraded from P.862 with only a little decrease in average correlation to listener opinions. And by comparing PESQ with GQI, the good performance of PESQ was verified.