发明名称 |
Method for classifying spoken language in spoken dialogue systems |
摘要 |
<p>The method involves applying a language recognition procedure and a language classification procedure on a digital language input for identifying a language e.g. German, of the language input by logic analysis of results of the procedures. The language input is digitally buffered, and additional digitized language data are determined from the language input. The language data are composed of feature extracted data and audio data, and are described by the language recognition procedure or the language classification procedure. Independent claims are also included for the following: (1) a device for classifying two languages in an automatic language dialogue system (2) a data carrier comprising instructions to perform a method for classifying two languages in an automatic language dialogue system.</p> |
申请公布号 |
EP1927980(A2) |
申请公布日期 |
2008.06.04 |
申请号 |
EP20070120606 |
申请日期 |
2007.11.13 |
申请人 |
DEUTSCHE TELEKOM AG |
发明人 |
ECKERT, MARTIN;ENGLERT, ROMAN;JOHANNSEN, WIEBKE;RUNGE, FRED;VAN BALLEGOOY, MARKUS |
分类号 |
G10L15/22;G10L15/00 |
主分类号 |
G10L15/22 |
代理机构 |
|
代理人 |
|
主权项 |
|
地址 |
|