Systems and methods for audio signal processing

Patent metadata

Field	Value
Publication number	US-9305567-B2
Application number	US-201313827894-A
Country	US
Kind code	B2
Filing date	Mar 14, 2013
Priority date	Apr 23, 2012
Publication date	Apr 5, 2016
Grant date	Apr 5, 2016

Abstract

A method for signal level matching by an electronic device is described. The method includes capturing a plurality of audio signals from a plurality of microphones. The method also includes determining a difference signal based on an inter-microphone subtraction. The difference signal includes multiple harmonics. The method also includes determining whether a harmonicity of the difference signal exceeds a harmonicity threshold. The method also includes preserving the harmonics to determine an envelope. The method further applies the envelope to a noise-suppressed signal.

First claim

What is claimed is:

1. A method for signal level matching of one or more audio signals by an electronic audio device, comprising:
segmenting an input spectrum into one or more bands;
measuring a signal-to-noise ratio for each band;
determining if the signal-to-noise ratios are less than a signal-to-noise ratio threshold;
assembling a target spectrum, wherein assembling a target spectrum comprises replacing a portion of a speech reference spectrum with a portion of a speech template spectrum that corresponds to one or more bands of the audio signals with signal-to-noise ratios that are less than the signal-to-noise ratio threshold; and
adjusting a gain of one or more bands in a noise-suppressed signal such that the one or more bands approximately match the target spectrum.

2. The method of claim 1 , further comprising:
capturing the audio signals from a plurality of microphones;
determining a difference signal based on an inter-microphone subtraction, wherein the difference signal comprises multiple harmonics;
determining whether a harmonicity of the difference signal exceeds a harmonicity threshold;
preserving the harmonics to determine an envelope; and
applying the envelope to the noise-suppressed signal.

3. The method of claim 2 , further comprising suppressing residual noise based on the audio signals.

4. The method of claim 2 , wherein applying the envelope to the noise-suppressed signal comprises adjusting the gain of the noise-suppressed signal such that a noise-suppressed signal level approximately matches an audio signal level.

5. The method of claim 2 , wherein determining a difference signal comprises determining portions of the input spectrum that correspond to a speech signal.

6. The method of claim 1 , wherein the speech template spectrum is based on a codebook.

7. The method of claim 1 , wherein the speech template spectrum is based on an interpolation of the bands of the input spectrum where the signal-to-noise ratio is greater than the signal-to-noise ratio threshold.

8. The method of claim 1 , wherein the speech reference spectrum is based on the input spectrum.

9. The method of claim 1 , wherein assembling a target spectrum comprises harmonic synthesis generation.

10. The method of claim 1 , further comprising:
generating a confidence measure or a voicing parameter;
estimating one or more sinusoidal parameters based on a fundamental frequency;
generating a sinusoidal signal based on the one or more sinusoidal parameters; and
multiplying the sinusoidal signal by the confidence measure or the voicing parameter to produce a scaled sinusoidal signal.

11. The method of claim 10 , further comprising:
filtering the scaled sinusoidal signal to produce a first filtered signal;
filtering the transformed signal to produce a second filtered signal;
summing the first filtered signal and the second filtered signal to produce a second summed signal; and
transforming the second summed signal into a time domain.

12. The method of claim 1 , wherein the target spectrum is based on gain differences and a pitch estimate.

13. The method of claim 1 , further comprising:
receiving a noise signal;
filtering the noise signal to produce a filtered noise signal;
generating a first summed signal based on the filtered noise signal and a speech signal;
generating a transformed signal based on the first summed signal;
generating a fundamental frequency of the transformed signal;
generating a confidence measure or a voicing parameter;
estimating one or more sinusoidal parameters based on the fundamental frequency;
generating a sinusoidal signal based on the one or more sinusoidal parameters;
multiplying the sinusoidal signal by the confidence measure or the voicing parameter to produce a scaled sinusoidal signal;
filtering the scaled sinusoidal signal to produce a first filtered signal;
filtering the transformed signal to produce a second filtered signal;
summing the first filtered signal and the second filtered signal to produce a second summed signal; and
transforming the second summed signal into a time domain.

14. An electronic audio device for signal level matching of one or more audio signals, comprising:
inter-microphone subtraction circuitry configured to segment an input spectrum into one or more bands;
peak sufficiency determination circuitry coupled to the inter-microphone subtraction circuitry, wherein the peak sufficiency determination circuitry is configured to measure a signal-to-noise ratio for each band and to determine if the signal-to-noise ratios are less than a signal-to-noise ratio threshold;
assemble spectrum circuitry coupled to the peak sufficiency determination circuitry, wherein the assemble spectrum circuitry is configured to assemble a target spectrum by replacing a portion of a speech reference spectrum with a portion of a speech template spectrum that corresponds to one or more bands of the audio signals with signal-to-noise ratios that are less than the signal-to-noise ratio threshold; and
a gain adjuster coupled to the assemble spectrum circuitry, wherein the gain adjuster is configured to adjust a gain of one or more bands in the noise-suppressed signal such that the one or more bands approximately match the target spectrum.

15. The electronic audio device of claim 14 , wherein the inter-microphone subtraction circuitry is configured to determine a difference signal based on an inter-microphone subtraction, wherein the difference signal comprises multiple harmonics, and wherein the electronic audio device further comprises:
a plurality of microphones coupled to the inter-microphone subtraction circuitry, wherein the plurality of microphones is configured to capture the audio signals;
envelope determination circuitry coupled to the inter-microphone subtraction circuitry, wherein the envelope determination circuitry is configured to determine whether a harmonicity of the difference signal exceeds a harmonicity threshold and to preserve the harmonics to determine an envelope; and
envelope application circuitry coupled to the envelope determination circuitry, wherein the envelope application circuitry is configured to apply the envelope to a noise-suppressed signal.

16. The electronic audio device of claim 15 , further comprising residual noise suppression circuitry coupled to the inter-microphone subtraction circuitry, wherein the residual noise suppression circuitry is configured to suppress residual noise based on the audio signals.

17. The electronic audio device of claim 15 , wherein the envelope application circuitry is configured to apply the envelope to the noise-suppressed signal by adjusting a gain of the noise-suppressed signal such that a noise-suppressed signal level approximately matches an audio signal level.

18. The electronic audio device of claim 15 , wherein the inter-microphone subtraction circuitry is configured to determine the difference signal by determining portions of the input spectrum that correspond to a speech signal.

19. The electronic audio device of claim 14 , wherein the speech template spectrum is based on a codebook.

20. The electronic audio device of claim 14 , wherein the speech template spectrum is based on an interpolation of the bands of the input spectrum where the signal-to-noise ratio is greater than the signal-to-noise ratio threshold.

21. The electronic audio device of claim 14 , wherein the speech reference spectrum is based on the input spectrum.

22. The electr

Assignees

• Qualcomm Inc

Inventors

• Visser Erik
• Kim Lae-Hoon
• Shin Jongwon
• Guo Yinyi
• Ryu Sang-Uk
• Schevciw Andre Gustavo P

Classifications

• G10L21/0316

  by changing the amplitude · CPC title

• G10L25/93

  Discriminating between voiced and unvoiced parts of speech signals (G10L25/90 takes precedence) · CPC title

• G10L15/20

  Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech (G10L21/02 takes precedence) · CPC title

• G10L2021/02165

  Two microphones, one receiving mainly the noise signal and the other one mainly the speech signal · CPC title

• G10L21/0208 Primary

  Noise filtering · CPC title