发明名称 Verfahren und Anordnung zur sprecherunabhängigen Spracherkennung für ein Telekommunikations- bzw. Datenendgerät
摘要 New names to enter in telephone book in phone are input using keypad and are transferred from telephone end unit (T,MS) via data transfer route (B) of communication or data network to central server (TS). Words are mapped to phonetic transcription. Phonetic transcription is transferred back via the data transfer route to the end unit (T,MS) and stored there. Number of named person is then selectable by voice input of the name. Input new names to add to telephone book are mapped to phonetic transcription using access to pronunciation library database (PDB1-3) and neural network implemented in server. Phonetic transcription is transferred back via the data transfer route to the end unit (T,MS) and stored in the end unit. Mapping to phonetic transcription is independent service of the telecommunication or data network. End unit identifier or user identifier is used to indicate e.g. dialect or pronunciation type for phonetic transcription. -
申请公布号 DE50003855(D1) 申请公布日期 2003.10.30
申请号 DE2000503855 申请日期 2000.12.18
申请人 SIEMENS AG 发明人 NIEMOELLER, MEINRAD
分类号 G10L15/06;G10L15/18;(IPC1-7):G10L15/26 主分类号 G10L15/06
代理机构 代理人
主权项
地址