发明名称 VOICE DETECTING SYSTEM
摘要 PROBLEM TO BE SOLVED: To detect distinctively a soundless and a sound-equipped division regardless of the type of noise included in an environment in which noises exist in the background, by detecting the noise during the soundless division of a voice and judging it to be soundless. SOLUTION: A digital voice signal fed from a terminal 101 is given to a signal subtraction part and also to a condition judging part 105 to make judgement if there is sound or not, and the result from condition judgement 108 is notified to the signal subtraction part 104. The soundless division notified according to the judging result 108 undergoes damping of its noise signal by the signal subtraction part 104, and the processed signal 111 is subjected to a parametic encoding by a voice encoding part 109 and also error correct encoding by an error correct encoding part 110, and the result is emitted from a terminal 102. Accordingly the condition judgement of sound-equipped division or soundless division can be made without increasing the power consumption or the program quantity of the software even when the fed digital voice signal includes noises in the background, and further, automatic updating of the threshold can be made, so that existence of a sound-equipped division can be sensed easily even with varying level of the background noise which is superposed over the digital voice signal.
申请公布号 JPH0934495(A) 申请公布日期 1997.02.07
申请号 JP19950185282 申请日期 1995.07.21
申请人 HITACHI LTD 发明人 SANO TSUTOMU
分类号 G10L11/02;(IPC1-7):G10L9/00 主分类号 G10L11/02
代理机构 代理人
主权项
地址