To provide a speech processing device for performing suppress processing effectively for transmitting a sound signal having a small amount of feeling of wrongness to the other party's device, and to provide a microphone apparatus to which the speech processing device is mounted.
An echo canceller 1 inputs a sound input signal for removing an echo component. A noise learning means 2 extracts a noise component from the sound signal from which echo is removed for calculating steady noise. A noise canceller 3 removes noise from the sound signal from which the echo is removed, based on the steady noise. An echo learning means 4 uses a sound output signal for learning residual echoes that are not removed by the echo canceller 1. An echo suppressor 5 removes the echo output from the noise canceller 3, and the residual echo from the sound signal from which the noise is removed. A speaker volume estimator 6 removes the steady noise and the residual echo from the sound signal from which the echo is removed for calculating estimation speaker volume. A suppress means 7 suppresses the voice signal according to the estimated speaker volume if required.
JP2003110690A | 2003-04-11 | |||
JP2003249996A | 2003-09-05 | |||
JP2003506924A | 2003-02-18 | |||
JP2005159677A | 2005-06-16 | |||
JPH09127982A | 1997-05-16 | |||
JP2003101445A | 2003-04-04 | |||
JPS62211698A | 1987-09-17 | |||
JP2004147069A | 2004-05-20 | |||
JPH07240703A | 1995-09-12 | |||
JP2004341339A | 2004-12-02 | |||
JP2007052150A | 2007-03-01 |