| 摘要 |
PROBLEM TO BE SOLVED: To transmit/receive, store, and reproduce a voice waveform so that a quantity of information and deterioration is minimized without limiting its pattern, and thereby to reduce communication costs. SOLUTION: A server 16 receives voice data transmitted from communication terminal equipment 13, and a feature value extraction part 20 of the server 16 extracts a feature value which is the information for identifying the phoneme of the voice from the received voice data, and a sound source frequency value which is the frequency value for producing the voice. These extracted feature value and sound source frequency value are stored in a data storage part 22. These stored feature value and sound source frequency value are read by a voice restoring part 24, and are restored to the original voice data by synthesizing the feature value and the frequency component corresponding to the sound source frequency value, and these restored voice data are transmitted to a communication terminal 14 from the server 16. COPYRIGHT: (C)2005,JPO&NCIPI
|
