发明名称 |
Method for speaker identification |
摘要 |
<p>To increase the robustness and/or the recognition rate for methods of speaker identification, it is proposed to include linguistic, paralinguistic, and/or extralinguistic speech features and, in particular, voice quality features to be extracted from a given speech input (SI) to identify a current speaker. <IMAGE></p> |
申请公布号 |
EP1280137(A1) |
申请公布日期 |
2003.01.29 |
申请号 |
EP20010117940 |
申请日期 |
2001.07.24 |
申请人 |
SONY INTERNATIONAL (EUROPE) GMBH |
发明人 |
MARASEK, KRZYSZTOF;TATO, RAQUEL;KEMP, THOMAS |
分类号 |
G10L17/10;(IPC1-7):G10L17/00 |
主分类号 |
G10L17/10 |
代理机构 |
|
代理人 |
|
主权项 |
|
地址 |
|