Noise suppression

Patent metadata

Field	Value
Publication number	US-10026415-B2
Application number	US-201515120130-A
Country	US
Kind code	B2
Filing date	Mar 2, 2015
Priority date	Mar 17, 2014
Publication date	Jul 17, 2018
Grant date	Jul 17, 2018

Abstract

A noise suppressor comprises a first (401) and a second transformer (403) for generating a first and second frequency domain signal from a frequency transform of a first and second microphone signal. A gain unit (405, 407, 409) determines time frequency tile gains in response to a difference measure for magnitude time frequency tile values of the first frequency domain signal and magnitude time frequency tile values of the second frequency domain signal. A scaler (411) generates a third frequency domain signal by scaling time frequency tile values of the first frequency domain signal by the time frequency tile gains; and the resulting signal is converted to the time domain by a third transformer (413). A designator (405, 407, 415) designates time frequency tiles of the first frequency domain signal as speech tiles or noise tiles; and the gain unit (409) determines the gains in response to the designation of the time frequency tiles as speech tiles or noise tiles.

First claim

The invention claimed is:

1. A noise suppressor for suppressing noise in a first microphone signal, the noise suppressor comprising:
a first transformer for generating a first frequency domain signal from a frequency transform of a first microphone signal, the first frequency domain signal being represented by time frequency tile values;
a second transformer for generating a second frequency domain signal from a frequency transform of a second microphone signal, the second frequency domain signal being represented by time frequency tile values;
a gain unit for determining time frequency tile gains as a non-negative monotonic function of a difference measure being indicative of a difference between a first monotonic function of a magnitude time frequency tile value of the first frequency domain signal and a second monotonic function of a magnitude time frequency tile value of the second frequency domain signal; and
a scaler for generating an output frequency domain signal by scaling time frequency tile values of the first frequency domain signal by the time frequency tile gains;
the noise suppressor further comprising:
a designator for designating time frequency tiles of the first frequency domain signal as speech tiles or noise tiles; and
the gain unit is arranged to determine the time frequency tile gains in response to the designation of the time frequency tiles of the first frequency domain signal as speech tiles or noise tiles such that a lower gain value for a time frequency tile gain of a time frequency tile is determined when the time frequency tile is designated as a noise tile than when the time frequency tile is designated as a speech tile.

2. The noise suppressor of claim 1 wherein the gain unit is arranged to determine a gain value for a time frequency tile gain of a time frequency tile as a function of the difference measure for the time frequency tile.

3. The noise suppressor of claim 2 wherein at least one of the first monotonic function and the second monotonic function is dependent on whether the time frequency tile is designated as a speech tile or as a noise tile.

4. The noise suppressor of claim 3 wherein the second monotonic function comprises a scaling of the magnitude time frequency tile value of the second frequency domain signal for the time frequency tile with a scale value dependent on whether the time frequency tile is designated as a speech time frequency tile or a noise time frequency tile.

5. The noise suppressor of claim 3 wherein the gain unit is arranged to generate a noise coherence estimate indicative of a correlation between an amplitude of the second microphone signal and an amplitude of a noise component of the first microphone signal and at least one of the first monotonic function and the second monotonic function is dependent on the noise coherence estimate.

6. The noise suppressor of claim 5 wherein the first monotonic function and the second monotonic function are such that an expected value of the difference measure is negative if an amplitude relationship between the first microphone signal and the second microphone signal corresponds to the noise coherence estimate and the time frequency tile is designated as a noise tile.

7. The noise suppressor of claim 6 wherein the gain unit is arranged to vary at least one of the first monotonic function and the second monotonic function such that the expected value of the difference measure for the amplitude relationship between the first microphone signal and the second microphone signal corresponding to the noise coherence estimate is different for a time frequency tile designated as a noise tile than for a time frequency tile designated as a speech tile.

8. The noise suppressor of claim 1 wherein the designator is arranged to designate time frequency tiles of the first frequency domain signal as speech tiles or noise tiles in response to difference values generated in response to the difference measure for a noise tile to the magnitude time frequency tile values of the first frequency domain signal and magnitude time frequency tile values of the second frequency domain signal.

9. The noise suppressor of claim 8 wherein the designator is arranged to filter difference values over a plurality of time frequency tiles, the filtering including time frequency tiles differing in both time and frequency.

10. The noise suppressor of claim 1 wherein the gain unit is arranged to filter gain values over a plurality of time frequency tiles, the filtering including time frequency tiles differing in both time and frequency.

11. The noise suppressor of claim 1 wherein the gain unit is arranged to filter at least one of the magnitude time frequency tile values of the first frequency domain signal and the magnitude time frequency tile values of the second frequency domain signal; the filtering including time frequency tiles differing in both time and frequency.

12. The noise suppressor of claim 1 further comprising an audio beamformer arranged to generate the first microphone signal and the second microphone signal from signals from a microphone array.

13. The noise suppressor of claim 1 further comprising an adaptive canceller for cancelling a signal component of the first microphone signal correlated with the second microphone signal from the first microphone signal.

14. A method of suppressing noise in a first microphone signal, the method comprising:
generating a first frequency domain signal from a frequency transform of a first microphone signal, the first frequency domain signal being represented by time frequency tile values;

generating a second frequency domain signal from a frequency transform of a second microphone signal, the second frequency domain signal being represented by time frequency tile values;
determining time frequency tile gains as a non-negative monotonic function of a difference measure being indicative of a difference between a first monotonic function of a magnitude time frequency tile value of the first frequency domain signal and a second monotonic function of a magnitude time frequency tile value of the second frequency domain signal; and
generating an output frequency domain signal by scaling time frequency tile values of the first frequency domain signal by the time frequency tile gains;
the method further comprising:
designating time frequency tiles of the first frequency domain signal as speech tiles or noise tiles; and wherein the time frequency tile gains are determined in response to the designation of the time frequency tiles of the first frequency domain signal as speech tiles or noise tiles such that a lower gain value for a time frequency tile gain of a time frequency tile is determined when the time frequency tile is designated as a noise tile than when the time frequency tile is designated as a speech tile.

15. A computer program product comprising computer program code means adapted to perform all the steps of claim 14 when said program is run on a computer.

Assignees

• Koninklijke Philips Nv

Inventors

• Janse Cornelis Pieternella
• Van Stuivenberg Leonardus Cornelis Antonius
• Kechichian Patrick

Classifications

• G10L21/0232 Primary

  Processing in the frequency domain · CPC title

• G10L21/0208 Primary

  Noise filtering · CPC title

• G10L2021/02166

  Microphone arrays; Beamforming · CPC title

• G10L21/04

  Time compression or expansion · CPC title

• G10L25/18

  the extracted parameters being spectral information of each sub-band · CPC title