发明名称 |
Method for determining a list of hypotheses from a vocabulary of a speech recognition system |
摘要 |
<p>The method involves entering a beginning of a word to be identified by individual letter sequence and by partial words (103) by a user. The identified letter sequences and the partial words are defragmented into a sequence of temporal input to an identified input. A distance dimension for a similarity between the identified input and entries of a vocabulary of a speech recognition system is determined, and a hypothesis list is represented with determined entries of the user on a display unit for selection.</p> |
申请公布号 |
EP2006835(A1) |
申请公布日期 |
2008.12.24 |
申请号 |
EP20080104216 |
申请日期 |
2008.06.02 |
申请人 |
SIEMENS AKTIENGESELLSCHAFT |
发明人 |
HEIDENREICH, SABINE |
分类号 |
G10L15/22;G01C21/36;G10L15/00;G10L15/04;G10L15/10;G10L15/18;G10L15/187;G10L15/26 |
主分类号 |
G10L15/22 |
代理机构 |
|
代理人 |
|
主权项 |
|
地址 |
|