Method and apparatus for wind noise detection and suppression using multiple microphones

• US 9,330,675 B2
• Filed: 09/30/2011
• Issued: 05/03/2016
• Est. Priority Date: 11/12/2010
• Status: Active Grant

• Current Assignee: Avago Technologies International Sales Pte Limited (Broadcom, Inc.)
• Original Assignee: Broadcom Corporation (Broadcom, Inc.)
• Inventors: Zhang, Xianxian, Chen, Juin-Hwey, Zeng, Huaiyu, Thyssen, Jes
• Primary Examiner(s): Lao, Lun-See
• Application Number:
  US13/250,355
• Publication Number:
  US 20120121100A1
• Time in Patent Office: 1,677 Days
• Field of Search: 381/92, 381 711- 718, 381 941- 944, 381/317, 704223-227, 455/296
• US Class Current: 1/1
• CPC Class Codes: G10L 2021/02165 Two microphones, one receiving mainly the noise signal and the other one mainly the speech signal
• CPC Class Codes: G10L 21/0208 Noise filtering
• CPC Class Codes: G10L 21/0272 Voice signal separating
• CPC Class Codes: H04R 1/245 of microphones
• CPC Class Codes: H04R 2410/07 Mechanical or electrical reduction of wind noise generated by wind passing a microphone

Abstract

Unlike sound based pressure waves that go everywhere, air turbulence caused by wind is usually a fairly local event. Therefore, in a system that utilizes two or more spatially separated microphones to pick up sound signals (e.g., speech), wind noise picked up by one of the microphones often will not be picked up (or at least not to the same extent) by the other microphone(s). Embodiments of methods and apparatuses that utilize this tact and others to effectively detect and suppress wind noise using multiple microphones that are spatially separated are described.

Claims

1. An apparatus for detecting and suppressing wind noise in a primary signal received by a primary microphone, the apparatus comprising:

a multi-method wind noise detection module configured to generate first and second wind noise detection signals that indicate whether wind noise is present or absent in the primary signal;

a wind noise detection signal combining module configured to combine the first and second wind noise detection signals to provide a primary microphone wind noise detection signal that indicates whether wind noise is present or absent in the primary signal; and

a primary microphone wind noise suppression module configured to determine a suppression gain for a sub-band of the primary signal based on;

the primary microphone wind noise detection signal, and

a comparison, performed by a suppression gain calculation module, of a difference in energy between the sub-band of the primary signal and a corresponding sub-band of a reference signal received by a reference microphone to a speech threshold and a wind threshold, wherein the speech threshold or the wind threshold is calculated based on the difference in energy and a previously calculated threshold value.

2. The apparatus of claim 1, wherein the primary microphone wind noise suppression module comprises:

an energy ratio calculation module configured to divide an estimate of the energy of the sub-band of the primary signal by an estimate of the energy of the corresponding sub-band of the reference signal to determine the difference in energy.

3. The apparatus of claim 1, wherein the primary microphone wind noise suppression module is further configured to determine the suppression gain such that the suppression gain is a constant value if the difference in energy is below the speech threshold or above the wind threshold.

4. The apparatus of claim 1, wherein the primary microphone wind noise suppression module is further configured to determine the suppression gain such that the suppression gain increases as the difference in energy increases from the speech threshold to the wind threshold.

5. The apparatus of claim 1, wherein the primary microphone wind noise suppression module is further configured to smooth the suppression gain over time.

6. The apparatus of claim 1, wherein the primary microphone wind noise suppression module further comprises:

a suppression gain mapping module configured to determine suppression gains for a group of frequency components in the sub-band of the primary signal by interpolating between the suppression gain and a suppression gain for an additional sub-band of the primary signal.

7. The apparatus of claim 1, wherein the first wind noise detection signal is generated based on the primary signal and the reference signal.

8. The apparatus of claim 1, wherein the multi-method wind noise detection module comprises:

a correlation based wind noise detection module configured to generate the first wind noise detection signal; and

a spectral deviation based wind noise detection module configured to generate the second wind noise detection signal.

9. The apparatus of claim 8, wherein the correlation based wind noise detection module is configured to generate the first wind noise detection signal based on:

the correlation of a first set of samples of the primary signal with a first set of samples of the reference signal,

the correlation of the first set of samples of the primary signal with a second set of samples of the primary signal, wherein the second set of samples of the primary signal are in a pitch period range of the first set of samples of the primary signal, and

the correlation of the first set of samples of the reference signal with a second set of samples of the reference signal, wherein the second set of samples of the reference signal are in a pitch period range of the first set of samples of the reference signal.

10. The apparatus of claim 9, wherein the spectral deviation based wind noise detection module is configured to generate the second wind noise detection signal by comparing a frequency spectrum of the primary microphone with a frequency spectrum associated with wind noise.

11. The apparatus of claim 1, further comprising a reference microphone wind noise suppression module.

12. A method for detecting and suppressing wind noise in a primary signal received by a primary microphone, the apparatus comprising:

generating first and second wind noise detection signals that indicate whether wind noise is present or absent in the primary signal;

combining the first and second wind noise detection signals to provide a primary microphone wind noise detection signal that indicates whether wind noise is present or absent in the primary signal; and

determining a suppression gain for a sub-band of the primary signal based on;

the primary microphone wind noise detection signal, and

a comparison, performed by a suppression gain calculation module, of a difference in energy between the sub-band of the primary signal and a corresponding sub-band of a reference signal received by a reference microphone to a speech threshold and a wind threshold, wherein the speech threshold or the wind threshold is calculated based on the difference in energy and a previously calculated threshold value.

13. The method of claim 12, where the determining the suppression gain further comprises:

dividing an estimate of the energy of the sub-band of the primary signal by an estimate of the energy of the corresponding sub-band of the reference signal to determine the difference in energy.

14. The method of claim 12, wherein the determining the suppression gain further comprises:

determining the suppression gain such that the suppression gain is a constant value if the difference in energy is below the speech threshold or above the wind threshold.

15. The method of claim 12, wherein the determining the suppression gain further comprises:

determining the suppression gain such that the suppression gain decreases as the difference in energy increases from the speech threshold to the wind threshold.

16. The method of claim 12, further comprising:

smoothing the suppression gain over time.

17. The method of claim 12, further comprising:

determining suppression gains for a group of frequency components in the sub band of the primary signal by interpolating between the suppression gain and a suppression gain for an additional sub-band of the primary signal.

18. The method of claim 12, wherein the generating the first wind noise detection signal comprises:

generating the first wind noise detection signal based on the primary signal and the reference signal.

19. The method of claim 12, wherein the generating the first wind noise detection signal comprises:

correlating a first set of samples of the primary signal with a first set of samples of the reference signal;

correlating the first set of samples of the primary signal with a second set of samples of the primary signal, wherein the second set of samples of the primary signal are in a pitch period range of the first set of samples of the primary signal; and

correlating, the first set of samples of the reference signal with a second set of samples of the reference signal, wherein the second set of samples of the reference signal are in a pitch period range of the first set of samples of the reference signal.

20. The method of claim 12, wherein generating the second wind noise detection signal comprises:

comparing a frequency spectrum of the primary microphone with a frequency spectrum associated with wind noise.