If the system determines the classification result is music, it is left open, otherwise go to the speaker identification stage. Identification prior to extraction and talk to themselves great mfcc characteristics, followed by vector quantization algorithm to generate a code word representing the ind

If the system judge of the classification result is music, that temporary do not handle, otherwise enter to talk the person to identify the stage.Before identify, first withdrew and talked the very big MFCC characteristic of person's relativity, immediately after turned the calculate way through an

If the system determination classified result is a music, that temporarily does not process, otherwise enters the storyteller to distinguish the stage.Before the recognition, first withdraws and the storyteller itself relevance very big MFCC characteristic, then through the vector quantification alg

If the system determine the classification of the results is the music, it is not processed or entered the stage of speaker recognition. Prior to recognition, extraction and correlation between a lot of the speaker itself MFCC feature, followed by vector quantization algorithm, for the training of e

正在翻譯，請(qǐng)等待...