发明名称 |
VOICE RECOGNITION SYSTEM, TERMINAL AND PROGRAM, AND VOICE RECOGNITION METHOD |
摘要 |
PROBLEM TO BE SOLVED: To provide a voice recognition system which is made suitable to improve recognition precision while conducting unspecified speaker voice recognition and to reduce a required memory capacity. SOLUTION: A voice model accumulating server 100 is provided with a voice model database 110 that stores a plurality of voice models in which specific voice parameters are made different for a same phoneme set and a voice model selection section 130 through which a plurality of voice models is selected among the database 110 based on voice specific information. A voice model selected by the section 130 is transmitted to a portable terminal 200. The terminal 200 is provided with a voice parameter extracting section 215 which is used to extract voice parameters from inputted voice being inputted by a microphone 205 and a voice recognition section 240 which is used to conduct voice recognition based on the voice parameters extracted by the section 215 and a voice model stored in a RAM 235. The voice model received from the server 100 is stored in the RAM 235. |
申请公布号 |
JP2003122395(A) |
申请公布日期 |
2003.04.25 |
申请号 |
JP20010322281 |
申请日期 |
2001.10.19 |
申请人 |
ASAHI KASEI CORP |
发明人 |
SHOSAKAI MAKOTO |
分类号 |
G10L15/28;G10L15/02;G10L15/06;(IPC1-7):G10L15/28 |
主分类号 |
G10L15/28 |
代理机构 |
|
代理人 |
|
主权项 |
|
地址 |
|