首页> 外国专利> CONTINUOUS WORD RECOGNITION SYSTEM FOR SPEECH RECOGNIZING DEVICE AND RECORDING MEDIUM STORED WITH CONTINUOUS WORD RECOGNITION PROCESSING PROGRAM IN SPEECH RECOGNIZING DEVICE

CONTINUOUS WORD RECOGNITION SYSTEM FOR SPEECH RECOGNIZING DEVICE AND RECORDING MEDIUM STORED WITH CONTINUOUS WORD RECOGNITION PROCESSING PROGRAM IN SPEECH RECOGNIZING DEVICE

机译:语音识别装置中的语音识别装置和存储有连续字识别处理程序的存储介质的连续字识别系统

摘要

PROBLEM TO BE SOLVED: To make it possible to efficiently and surely define the recognition results of all words constituting continuous words by outputting (m) pieces of predetermined candidates sequentially from the first place, inputting the response indicating the correctness or not from a speaker by each outputted candidate, defining the response as the word to be recognized if the response is correct, outputting the next candidate if the response is no and executing this processing up to the m-th candidate. ;SOLUTION: Predetermined (m) pieces of the candidates are successively outputted from the first order (S56) and the response indicating the correctness or not from the speaker is inputted by each of the outputted candidates and is subjected to recognition processing. When the response of the speaker is decided to be correct (S68), this candidate is defined as the word to be recognized (S69). When the response is decided to be no (S71), the next candidate is outputted and the response indicating the correctness or not from the speaker is inputted to the outputted candidates and is subjected to the recognition processing. When the response of the speaker is decided to be correct, this candidate is defined as the word to be recognized (S70). If the response is no, the processing to output the next candidate is executed to the n-th candidate (S73).;COPYRIGHT: (C)2000,JPO
机译:解决的问题:通过从头开始依次输出(m)条预定候选,并从说话者输入指示正确与否的响应,从而有可能有效,可靠地定义构成连续单词的所有单词的识别结果。每个输出的候选者,如果响应正确,则将响应定义为要识别的单词,如果响应为否,则输出下一个候选者,并执行此处理直到第m个候选者。 ;解决方案:从一阶顺序地输出预定(m)个候选者(S56),并且由每个输出的候选者输入表示来自说话者的正确与否的响应,并进行识别处理。当确定讲话者的响应正确时(S68),将该候选者定义为要识别的单词(S69)。当确定响应为否(S71)时,输出下一个候选者,并且将来自说话者的指示正确与否的响应输入至输出的候选者,并进行识别处理。当确定讲话者的响应正确时,将该候选者定义为要识别的单词(S70)。如果响应为否,则将输出下一个候选者的处理执行到第n个候选者(S73)。COPYRIGHT:(C)2000,JPO

著录项

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号