摘要 |
<p><P>PROBLEM TO BE SOLVED: To provide a highly convenient speech retrieval device and a speech distribution system capable of referring to conference contents in parallel even during the progress of a speech conference and efficiently specifying a speech place necessary for a user. <P>SOLUTION: The speech retrieval device 101 detects the presence/absence of a speech from a speech acquired by a speech conference device 102, and detects a speaker name from a speech acquisition position, and detects words spoken from a speech waveform, and generates a speech file by discriminating speechs each time the presence of a speech changes, and distributes each speech file to a network storage device 103 for storage. The speech retrieval device 101 records the speaker name and the spoken words as the index of each speech file, and makes a client 104 display a list in which speech files retrieved based on this index are time-sequentially arranged, and distributes the speech files selected from the list through a network in parallel from a network storage device 103. <P>COPYRIGHT: (C)2008,JPO&INPIT</p> |