The Study Of VQ Speaker Recognition

The speaker recognition is the processing of automatically recognition whether the speaker in the speakers which records gathers, then determined who the speaker is, by analyzing the speaker's pronunciation signals and picking up the speaker's characteristic. With the development of computer technology and information based society, the speaker recognition technology receives more and more attention.Currently in the field of speaker recognition, there are two important questions need to solve for the enhancement of recognition rate. One is how to select more effective and more reliable speaker characteristic, the other is how to select the best distorted measure. This article has made the discussion in light of this, and made following aspects improvement and research.1. This article introduces a new algorithm that could better describe the speaker voice characteristic's fine structure in the characteristic space. And use this algorithm in the WDMVQ system. And in the article, we has done massive contrasts experiment.2. In the computation of weighting coefficient we use the empirical value. And this article redefines the function so that we could expect the weighting coefficient may be able to accurately reflect the divergence leaves of each component. And has used the theory of the importance of the MFCC in the discussion of the empirical value. And we carry on the discussion of whether to discard the second component of MFCC. The experiment results prove that WDMVQ with new weighting coefficient is better.