System and method for suppression of non-linear acoustic echoes
A method for residual echo suppression is provided. Embodiments may include receiving an original reference signal and applying a distortion function to the original reference signal to generate a second signal. Embodiments may include generating a non-linear signal from the distortion function that does not include linear components of the original reference signal. Embodiments may also include calculating a residual echo power of a linear component and a non-linear component, wherein the linear component is based upon the original reference signal and the non-linear component is based upon the non-linear signal. Embodiments may further include applying a room model to each of the original reference signal and the non-linear signal and estimating a power associated with the original reference signal and the non-linear signal. Embodiments may include calculating a combined echo power estimate as a weighted sum of a weighted original reference signal power and a weighted non-linear signal power.
1. A method for residual echo suppression comprising:
receiving, at a processor, an original reference signal and applying a distortion function to the original reference signal to generate a second signal;
generating a non-linear signal from the distortion function that does not include linear components of the original reference signal;
calculating a residual echo power of a linear component and a non-linear component, wherein the linear component is based upon, at least in part, the original reference signal and the non-linear component is based upon, at least in part, the non-linear signal;
applying a room model to each of the original reference signal and the non-linear signal;
estimating a power associated with the original reference signal and the non-linear signal; and
calculating a combined echo power estimate as a weighted sum of a weighted original reference signal power and a weighted non-linear signal power.
2. The method for residual echo suppression of claim 1 , wherein the application of the room model includes convolving each of the original reference signal and the non-linear signal with a room impulse response.
3. The method for residual echo suppression of claim 2 , wherein the room impulse response is estimated using a linear echo canceller.
4. The method for residual echo suppression of claim 1 , wherein at least one of the weighted original reference signal power and the weighted non-linear signal power used in the weighted sum include one or more coupling factors.
5. The method for residual echo suppression of claim 4 , wherein, the coupling factors differ between the linear component and the non-linear component.
6. The method for residual echo suppression of claim 1 , wherein a magnitude is used instead of the power.
7. The method for residual echo suppression of claim 4 , wherein the distortion function is estimated based upon, at least in part, a coupling factor associated with the non-linear signal.
8. The method for residual echo suppression of claim 7 , wherein the coupling factor includes at least one of broadband or an individual frequency band calculation.
9. The method for residual echo suppression of claim 1 , wherein generating a non-linear signal includes subtracting the original reference signal from the second signal to generate a non-linear signal.
10. The method for residual echo suppression of claim 1 , wherein the distortion function is approximated by a polynomial and wherein one or more coefficients of the polynomial are estimated based upon, at least in part, a correlation between a plurality of monomials and a cross-correlation of each of the plurality of monomials with the original reference signal.
11. The system for residual echo suppression of claim 1 , wherein generating a non-linear signal includes subtracting the original reference signal from the second signal to generate a non-linear signal.
12. The system for residual echo suppression of claim 1 , wherein the distortion function is approximated by a polynomial and wherein one or more coefficients of the polynomial are estimated based upon, at least in part, a correlation between a plurality of monomials and a cross-correlation of each of the plurality of monomials with the original reference signal.
13. A system for residual echo suppression comprising:
a speaker;
a microphone; and
a processor configured to perform one or more operations including:
receiving, at a processor, an original reference signal and applying a distortion function to the original reference signal to generate a second signal;
generating a non-linear signal from the distortion function that does not include linear components of the original reference signal;
calculating a residual echo power of a linear component and a non-linear component, wherein the linear component is based upon, at least in part, the original reference signal and the non-linear component is based upon, at least in part, the non-linear signal;
applying a room model to each of the original reference signal and the non-linear signal;
estimating a power associated with the original reference signal and the non-linear signal; and
calculating a combined echo power estimate as a weighted sum of a weighted original reference signal power and a weighted non-linear signal power.
14. The system for residual echo suppression of claim 13 , wherein the application of the room model includes convolving each of the original reference signal and the non-linear signal with a room impulse response.
15. The system for residual echo suppression of claim 14 , wherein the room impulse response is estimated using a linear echo canceller.
16. The system for residual echo suppression of claim 13 , wherein at least one of the weighted original reference signal power and the weighted non-linear signal power used in the weighted sum include one or more coupling factors.
17. The system for residual echo suppression of claim 16 , wherein, the coupling factors differ between the linear component and the non-linear component.
18. The system for residual echo suppression of claim 13 , wherein a magnitude is used instead of the power.
19. The system for residual echo suppression of claim 16 , wherein the distortion function is estimated based upon, at least in part, a coupling factor associated with the non-linear signal.
20. The system for residual echo suppression of claim 19 , wherein the coupling factor includes at least one of broadband or an individual frequency band calculation.