发明名称 |
SPEECH RECOGNITION SYSTEM, SPEECH RECOGNITION METHOD, AND SPEECH RECOGNITION PROGRAM |
摘要 |
<p>A speech recognition system and method capable of providing the result of speech recognition appropriate for each of a plurality of applications. A common data processing means (210) recognizes a speech captured by a speech input means (100) to generate a hypothesis of the result of the recognition, which has no bias in each application. An adaptive data processing means (220) regenerates the hypothesis of the recognition result by using adaptive data or an adaptive processing for each application to provide the result of the recognition recalculated for each application to the applications.</p> |
申请公布号 |
WO2007108500(A1) |
申请公布日期 |
2007.09.27 |
申请号 |
WO2007JP55811 |
申请日期 |
2007.03.22 |
申请人 |
NEC CORPORATION;YAMAMOTO, HITOSHI |
发明人 |
YAMAMOTO, HITOSHI |
分类号 |
G10L15/18;G10L15/22 |
主分类号 |
G10L15/18 |
代理机构 |
|
代理人 |
|
主权项 |
|
地址 |
|