Improving of a GMM Based Voice Conversion Method

碩士 === 國立臺灣科技大學 === 資訊工程系 === 98 === In this thesis, a GMM based voice conversion method is improved. The voice of a source speaker is converted to a target-speaker timbre liking voice while the content of the voice need not be known in advance. We have proposed three ways to improve the over smooth...

Full description

Bibliographic Details
Main Authors: Sung-feng Tsai, 蔡松峯
Other Authors: Hung-Yan Gu
Format: Others
Language:zh-TW
Published: 2009
Online Access:http://ndltd.ncl.edu.tw/handle/62254558219384307474