发明名称 SPEECH PROCESSING APPARATUS AND PROGRAM
摘要 PROBLEM TO BE SOLVED: To generate natural output speech from input speech as regards technology to change the characteristics of the speech. SOLUTION: An envelope specification section 23 generates input envelope data DEVin indicating the spectrum envelope EVin of an input speech signal Sin. A template acquisition section 33 reads spectrum data DSPt for conversion indicating the frequency spectrum SPt of speech for conversion out of a memory section 51. A data generation section 3a specifies a frequency spectrum SPnew which is the frequency spectrum of a shape corresponding to the frequency spectrum SPt of the speech for conversion and has the spectrum envelope which nearly coinciding with the spectrum envelope EVin of the input speech on the basis of the input envelope data DEVin and the spectrum data DSPt for conversion and generates new spectrum data DSPnew indicating the frequency spectrum SPnew. A reverse FFT section 15 and an output processing section 16 generate an output speech signal Snew on the basis of the new spectrum data DSPnew. COPYRIGHT: (C)2006,JPO&NCIPI
申请公布号 JP2006017946(A) 申请公布日期 2006.01.19
申请号 JP20040194800 申请日期 2004.06.30
申请人 YAMAHA CORP 发明人 YOSHIOKA YASUO;ALEX ROSUKOSU
分类号 G10L13/00;G10L13/02;G10L21/00;G10L21/04 主分类号 G10L13/00
代理机构 代理人
主权项
地址