摘要 |
PURPOSE:To eliminate a trouble to listen to the opinion of a speaker and write it in a blackboard and a time loss as in the conventional manner and to eliminate the omission of writing by displaying the voice signal of the speaker by means of automatic recognition and of making it into a sentence. CONSTITUTION:When a voice recognition device 4 receives a voice recording start signal from a voice recording device 3, it reads the recorded voice signal per channel, converts it into an analogue signal. Then, a acoustic feature spectrum is extracted by frequency-analyzing the signal at every constant period, and the voice signal is phoneme-analyzed by using a voice neuro-network and the like, whereby it is segmented in a terms of words or blocks. A sentence preparation device 5 makes the words or block data from the device 4 into the sentence by referring to a sentence preparation dictionary. An electronic blackboard main body 6 receives sentence data from the device 5 per channel together with the name of the speaker, displays the sentence in a display part 6A and stores it in a memory part 6B, whereby it executes copying a copying part 6C apart from a conventional copying function. |