发明名称 |
A VOICE-BASED GENDER IDENTIFICATION METHOD USING A PITCH FEATURE VECTOR AND A SUPPORT VECTOR MACHINE(SVM) |
摘要 |
A voice-based gender identification method using a pitch feature vector and a support vector machine is provided to significantly enhance a gender recognition performance in comparison with a conventional gender identification method using GMM(Gaussian Mixture Model) by using a combination feature vector and applying the support vector machine to voice-based gender identification. A voice-based gender identification method using a pitch feature vector and a support vector machine includes: detecting a voice from a voice database storing male's voice and female's voice; extracting a MFCC(Mel Frequency Cepstral Coefficient) feature vector for training from the detected voice; extracting a training pitch feature vector from the detected voice; combining the training pitch feature vector with the extracted training MFCC feature vector to produce a combination feature vector; and obtaining an optimal weight vector and an optimal bias based on the produced combination feature vector.
|
申请公布号 |
KR20080077880(A) |
申请公布日期 |
2008.08.26 |
申请号 |
KR20070017634 |
申请日期 |
2007.02.21 |
申请人 |
INHA-INDUSTRY PARTNERSHIP INSTITUTE |
发明人 |
CHANG, JOON HYUK;LEE, KYE HWAN;KIM, DEOK HWAN |
分类号 |
G10L15/02;G10L15/28;G10L25/51;G10L25/90 |
主分类号 |
G10L15/02 |
代理机构 |
|
代理人 |
|
主权项 |
|
地址 |
|