发明名称 SPEECH PROCESSING UNIT, SPEECH PROCESSING SYSTEM, AND SPEECH PROCESSING METHOD
摘要 PROBLEM TO BE SOLVED: To make it easier to see a mistake of a speech recognition result.SOLUTION: A speech processing unit 100 includes a speech recognition part 110, a selection part 120, and a response speech generation part 130. The speech recognition part 110 extracts a plurality of word candidates from a speech. The selection part 120 selects a different part among each of the word candidates based on the plurality of word candidates recognized by the speech recognition part 110. The response speech generation part 130 outputs a response speech in which the different part is made louder among the word candidates.
申请公布号 JP2014194480(A) 申请公布日期 2014.10.09
申请号 JP20130070682 申请日期 2013.03.28
申请人 FUJITSU LTD 发明人 TOGAWA TARO;ISHIKAWA CHISATO;OTANI TAKESHI
分类号 G10L15/22 主分类号 G10L15/22
代理机构 代理人
主权项
地址