摘要 |
<p>PURPOSE:To rapidly search for the necessary portion from a recording medium in which voice information is stored by deciding whether the voice keywords inputted from a microphone and the voice reproduced signals from a voice recording and reproducing device are coincident or not. CONSTITUTION:The voice reproduced data from a voice recording and reproducing device 1 are coded through a voice coding device 2 and inputted to an unspecified speaker voice recognition device 3. On the other hand, voice data are outputted to the device 2 as voice keywords through a voice input device 4 such as a microphone. Based on the coded voice keywords, the means 3 successively sends out voice reproduced data from the device 1 until coded voice reproduced data practically coincide and continues the retrieval.</p> |