Hearing perception characteristic-based objective voice quality evaluation method

Abstract

The invention discloses a hearing perception characteristic-based objective voice quality evaluation method which is simple and effective. An ear hearing model and non-linear compression conversion are introduced into an extraction process of MFCC (Mel frequency cepstrum coefficient) characteristic parameters according to psychoacoustics principles. According to the method, a Gammatone filter is adopted to simulate a cochlea basement membrane; and the strength-loudness perception characteristics of the voice are simulated through cube root non-linear compression conversion in an amplitude non-linear conversion process. By using new characteristic parameters, a voice quality evaluation method which is more accordant with the ear hearing perception characteristics is provided. Compared with other methods, the relevancy between objective evaluation results and subjective evaluation results is effective improved, the operation time is shorter and the complexity is lower, and the method has stronger adaptability, reliability and practicability. A new solution to improve the objective voice quality evaluation can be provided through the method for voice quality evaluation by simulating the hearing perception characteristics of human ears.

Description

technical field [0001] The present invention relates to an objective speech quality evaluation technology based on the auditory perception characteristics of the human ear, and more specifically, to a method of introducing the auditory model of the human ear into the extraction process of MFCC characteristic parameters, and realizing speech quality by calculating the degree of distortion of the characteristic parameters. Methods for objective quality assessment. Background technique [0002] Voice quality evaluation is one of the basic standards to measure the performance of voice communication system. From the perspective of the evaluation subject, it can be divided into two categories: subjective evaluation and objective evaluation. The MOS (Mean Opinion Score) method proposed by ITU-T Recommendation P.830 is a widely used subjective evaluation method, which uses the average opinion score of testers to intuitively reflect people’s perception of voice quality, but such ...

AI Technical Summary

Problems solved by technology

[0003] The PESQ perceptual speech quality evaluation proposed by ITU-T recommendation P.862 is currently a high-performance objective speech quality evaluation method, which can better identify communication delays, environmental noise and errors, but it is based on the perception model of the Bark spectrum. High computational complexity, which is not conducive to real-time evaluation of voice quality

[0004] The Mel-CD distortion measurement uses MFCC as the speech feature parameter, and has low computational complexity. It is a simple and effective speech quality evaluation method, but its evaluation performance is quite different from that of PESQ.

The analysis shows that although the auditory principle of the human ear and the decorrelation characteristics of Mel cepstrum are used in the extraction process of MFCC feature parameters, it uses a triangular filter bank to simulate the frequency selection characteristics of the cochlear basilar membrane and a logarithmic operation to simulate the amplitude The value nonlinear transformation process does not fully reflect the auditory perception characteristics of the human ear

Images