Multichannel acoustic echo reduction

• US 9,264,807 B2
• Filed: 01/23/2013
• Issued: 02/16/2016
• Est. Priority Date: 06/19/2008
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Inventors: Tashev, Ivan Jelev, Acero, Alejandro, Madhu, Nilesh
• Primary Examiner(s): Paul, Disler
• Application Number:
  US13/747,535
• Publication Number:
  US 20130129101A1
• Time in Patent Office: 1,119 Days
• Field of Search: 381/83, 381/93, 381/66, 37940601-40609, 37940611-40616
• US Class Current: 1/1
• CPC Class Codes: H04M 9/082 using echo cancellers echo cancellers per se H04B3/23
• CPC Class Codes: H04R 3/005 for combining the signals of two or more microphones specially adapted for hearing aids H04R25/407
• CPC Class Codes: H04R 3/02 for preventing acoustic reaction , i.e. acoustic oscillatory feedback specially adapted for hearing aids H04R25/453

Abstract

A multichannel acoustic echo reduction system is described herein. The system includes an acoustic echo canceller (AEC) component having a fixed filter for each respective combination of loudspeaker and microphone signals and having an adaptive filter for each microphone signal. For each microphone signal, the AEC component modifies the microphone signal to reduce contributions from the outputs of the loudspeakers based at least in part on the respective adaptive filter associated with the microphone signal and the set of fixed filters associated with the respective microphone signal.

Claims

1. A method executed by a processor of a computing device, the method comprising:

transmitting a first calibration signal to a first speaker, the first calibration signal causing the first speaker to generate first output;

transmitting a second calibration signal to a second speaker subsequent to transmitting the first calibration signal to the first speaker, the second signal causing the second speaker to generate second output;

receiving a first microphone signal from a first microphone, the first microphone signal corresponding to the first output from the first speaker;

receiving a second microphone signal from the first microphone, the second microphone signal corresponding to the second output from the second speaker;

computing a first coefficient of a first fixed filter for the first microphone based upon the first calibration signal and the first microphone signal;

computing a second coefficient of a second fixed filter for the first microphone based upon the second calibration signal and the second microphone signal; and

subsequent to computing the first coefficient of the first fixed filter and the second coefficient of the second fixed filter, applying the first fixed filter and the second fixed filter to an acoustic signal captured by the first microphone.

2. The method of claim 1, wherein applying the first fixed filter to the acoustic signal causes a first filter output to be generated, wherein applying the second fixed filter to the acoustic signal causes a second filter output to be generated, the method further comprising:

combining the first filter output with the second filter output to generate combined output.

3. The method of claim 2, further comprising applying a first adaptive filter over the combined output to generate a first filtered output.

4. The method of claim 3, further comprising updating at least one configuration value of the first adaptive filter based at least in part upon the acoustic signal captured by the first microphone and previous configuration values of the first adaptive filter.

5. The method of claim 3, further comprising subtracting the first filtered output from the acoustic signal to generate a first filtered microphone signal.

6. The method of claim 1, wherein the transmitting of the first calibration signal and the transmitting of the second calibration signal is undertaken responsive to the computing device being started.

7. The method of claim 1, wherein the acoustic signal comprises output from the first speaker, output from the second speaker, and audible output from a human, and wherein the first fixed filter and the second fixed filter facilitate filtering of the output from the first speaker and the output from the second speaker from the acoustic signal.

8. The method of claim 1, wherein the first output and the second output are respective acoustic signals.

9. The method of claim 1, wherein the acoustic signal comprises stereo output from the first speaker and the second speaker, and wherein the first fixed filter is employed to filter a first contribution of the first speaker to the stereo output and the second fixed filter is employed to filter a second contribution of the second speaker to the stereo output.

10. The method of claim 1, further comprising:

receiving a third microphone signal from a second microphone, the third microphone signal corresponding to the first output from the first speaker over the first time period;

receiving a fourth microphone signal from the second microphone, the fourth microphone signal corresponding to the second output from the second speaker over the second time period;

computing a third coefficient of a third fixed filter for the second microphone based upon the first calibration signal and the third microphone signal;

computing a fourth coefficient of a fourth fixed filter for the second microphone based upon the second calibration signal and the fourth microphone signal; and

subsequent to computing the third coefficient of the third fixed filter and the fourth coefficient of the fourth fixed filter, applying the third fixed filter and the fourth fixed filter to the acoustic signal captured by the second microphone.

11. A method, comprising:

transmitting a calibration signal to a first speaker and a second speaker to cause the first speaker to output a first speaker signal over a first time period and the second speaker to output a second speaker signal over a second time period, the second time period being subsequent to the first time period;

receiving, from a microphone, a first microphone signal for the first time period and a second microphone signal for the second time period; and

computing coefficients for a plurality of fixed filters for the microphone based at least in part upon the calibration signal, the first microphone signal, and the second microphone signal, the plurality of fixed filters facilitating filtering of signals output by the first speaker and the second speaker from an acoustic signal captured by the microphone.

12. The method of claim 11, the acoustic signal comprising spoken words from a human not included in the signals output by the first speaker and the second speaker, the method further comprising recognizing at least one word in the spoken words from the human based upon the filtering of the signals output by the first speaker and the second speaker.

13. The method of claim 11, further comprising updating an adaptive filter based at least in part upon the acoustic signal, wherein the adaptive filter is applied to a combination of outputs of the plurality of fixed filters.

14. The method of claim 13, further comprising applying a beamformer over output of the adaptive filter.

15. The method of claim 11 executed on a computing device that is executing voice recognition software.

16. The method of claim 11, further comprising computing multiple fixed filters for a plurality of microphones based upon respective signals received from the plurality of microphones during the first time period and the second time period.

17. The method of claim 11, further comprising updating a configuration of an adaptive filter for the microphone based at least in part upon the acoustic signal captured by the microphone.

18. The method of claim 11 configured for execution by a speakerphone.

19. The method of claim 11, wherein the first speaker signal and the second speaker signal are respective acoustic signals.

20. A computing device that is configured to perform an action responsive to receipt of a voice command, the computing device comprising a processor and a computer-readable medium that includes instructions that, when executed by the processor, causes the processor to perform acts comprising:

setting first coefficients for a first fixed filter for a microphone, wherein the first coefficients are set based upon a first output signal output by a first speaker and captured by the microphone at a first time, the first output signal comprises a calibration signal;

setting second coefficients for a second fixed filter for the microphone, wherein the second coefficients are set based upon a second output signal output by a second speaker and captured by the microphone at a second time, wherein the second time is subsequent the first time; and

subsequent to the first coefficients being set for the first filter and the second coefficients being set for the second filter, using the first fixed filter and the second fixed filter to modify a signal captured by the microphone.