BEAMFORMING PRE-PROCESSING FOR SPEAKER LOCALIZATION

• US 20140153740A1
• Filed: 02/10/2014
• Published: 06/05/2014
• Est. Priority Date: 07/16/2008
• Status: Active Grant

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Wolff, Tobias, Buck, Markus, Schmidt, Gerhard Uwe
• Granted Patent: US 9,414,159 B2
• Time in Patent Office: Days
• US Class Current: 381/92
• CPC Class Codes: H04R 2430/20 Processing of the output signals of the acoustic transducers of an array for obtaining a desired directivity characteristic H04R2203/12 takes precedence
• CPC Class Codes: H04R 2430/23 Direction finding using a sum-delay beam-former
• CPC Class Codes: H04R 3/005 for combining the signals of two or more microphones specially adapted for hearing aids H04R25/407

Abstract

Methods and apparatus to beamform a first plurality of microphone signals using at least one beamforming weight to obtain a first beamformed signal, beamform a second plurality of microphone signals using the at least one beamforming weight to obtain a second beamformed signal, and adjust the at least one beam forming weight so that the power density of at least one perturbation component present in the first or the second plurality of microphone signals is reduced.

Claims

1. A method comprising:

beamforming a first plurality of microphone signals using at least one beamforming weight to obtain a first beamformed signal;

beamforming a second plurality of microphone signals using the at least one beamforming weight to obtain a second beamformed signal; and

adjusting the at least one beamforming weight so that the power density of at least one perturbation component present in the first or the second plurality of microphone signals is reduced.

2. The method according to claim 1, further comprising receiving the first plurality of microphone signals from a first microphone array and receiving the second plurality of microphone signals from a second microphone array, wherein the first microphone array and the second microphone array share at least one microphone.

3. The method according to claim 1, wherein the at least one perturbation component comprises a noise component.

4. The method according to claim 1, wherein the at least one perturbation component comprises an echo component.

5. The method according to claim 1, further including performing speaker localization from the first and/or second microphone signals.

6. The method according to claim 5, further including determining a direction of the speaker with respect to a first microphone array providing the first microphone signals and/or a second microphone array providing the second microphone signals.

7. The method according to claim 5, further including determining a distance of the speaker with respect to a first microphone array providing the first microphone signals and/or a second microphone array providing the second microphone signals.

8. The method according to claim 1, further including performing calibration with no wanted signal present, wherein a wanted signal corresponds to a speech utterance.

9. The method according to claim 8, further including performing the calibration prior to the speech utterance being present.

10. The method according to claim 8, further including performing the calibration during speech pauses.

15. The article according to claim 1, further including instructions to perform speaker localization from the first and/or second microphone signals.

16. The article according to claim 15, further including instructions to determine a direction of the speaker with respect to a first microphone array providing the first microphone signals and/or a second microphone array providing the second microphone signals.

17. The article according to claim 15, further including instructions to determine a distance of the speaker with respect to a first microphone array providing the first microphone signals and/or a second microphone array providing the second microphone signals.

18. The article according to claim 15, further including instructions to perform calibration with no wanted signal present, wherein a wanted signal corresponds to a speech utterance.

19. The article according to claim 18, further including instructions to perform the calibration prior to the speech utterance being present.

11. An article, comprising:

a non-transitory computer readable storage medium having stored instructions that enable a machine to;

beamform a first plurality of microphone signals using at least one beamforming weight to obtain a first beamformed signal;

beamform a second plurality of microphone signals using the at least one beamforming weight to obtain a second beamformed signal; and

adjust the at least one beamforming weight so that the power density of at least one perturbation component present in the first or the second plurality of microphone signals is reduced.

12. The article according to claim 11, further comprising instructions to receive the first plurality of microphone signals from a first microphone array and receive the second plurality of microphone signals from a second microphone array, wherein the first microphone array and the second microphone array share at least one microphone.

13. The article according to claim 11, wherein the at least one perturbation component comprises a noise component.

14. The article according to claim 11, wherein the at least one perturbation component comprises an echo component.

20. A system, comprising:

a first beamformer module to receive a first plurality of microphone signals using at least one beamforming weight to obtain a first beamformed signal;

a second beamformer module to receive a second plurality of microphone signals using the at least one beamforming weight to obtain a second beamformed signal; and

a control module, which includes a processor, to adjust the at least one beamforming weight so that the power density of at least one perturbation component present in the first or the second plurality of microphone signals is reduced.