摘要 |
PROBLEM TO BE SOLVED: To provide a system for accurately eliminating signal components (noise components), except for the target voice from among a plurality of signals in which a plurality of voices and noise are mixed, and to provide a method therefor. SOLUTION: The system separates an input time series signal from a signal separating part 1, estimates the noise components included in a separation signal in a noise estimating part 2, by using a plurality of the separating signals, and eliminates noise estimated from the separation signal in a noise eliminating part 4. COPYRIGHT: (C)2006,JPO&NCIPI
|