| 摘要 |
An embodiment provides a sound source separation device comprising: a voice signal blocking unit which divides a sound source signal by one second intervals and performs a block process of the same when the sound source signal including a voice signal and a noise signal is inputted; a noise removal unit which performs a blind convolution process based on the sound source signal and outputs the voice signal in which the noise signal is removed; an impulse response obtaining unit which obtains an impulse response of the noise removal unit based on the voice signal; and a phase compensating unit which compensates for phase distortion of the voice signal by applying the impulse response in a reverse direction for the time. The sound source separation device can compensate for phase distortion of a voice signal in a process which removes a noise signal in a sound source signal and separates the voice signal. |
