Estimation of background noise in audio signals

Patent metadata

Field	Value
Publication number	US-11164590-B2
Application number	US-202016738504-A
Country	US
Kind code	B2
Filing date	Jan 9, 2020
Priority date	Dec 19, 2013
Publication date	Nov 2, 2021
Grant date	Nov 2, 2021

Abstract

The disclosure relates to a background noise estimator and a method therein, for supporting sound activity detection in an audio signal segment. The method comprises reducing a current background noise estimate when the audio signal segment is determined to comprise music and the current background noise estimate exceeds a minimum value. This is to be performed when an energy level of an audio signal segment is more than a threshold higher than a long term minimum energy level, lt_min, which is determined over a plurality of preceding audio signal segments, or, when the energy level of the audio signal segment is less than a threshold higher than lt_min, but no pause is detected in the audio signal segment.

First claim

The invention claimed is:

1. A method by an apparatus, the method comprising:
performing by at least one processor of the apparatus:
when a current energy level of an audio signal segment is higher than a current sub-band noise estimate and no pause is in the audio signal segment, reducing the current sub-band noise estimate by reducing the current sub-band noise estimate by a defined amount for at least one sub-band of the audio signal segment.

2. The method according to claim 1 , wherein reducing the current sub-band noise estimate when the current energy level of the audio signal segment is higher than the current sub-band noise estimate and no pause is in the audio signal segment comprises reducing the current sub-band noise estimate by the defined amount for the at least one sub-band of the audio signal segment responsive to the audio signal segment comprising music when the current energy level of the audio signal segment is higher than the current sub-band noise estimate, and no pause is in the audio signal segment.

3. The method according to claim 1 , wherein no pause is in the audio signal segment when one or both of the following is fulfilled:
a predefined number of consecutive preceding audio signal segments have been determined to comprise an active signal;
a dynamic of an audio signal comprising an audio signal segment does not exceed a signal dynamics threshold.

4. The method according to claim 1 , further comprising responsive to the current sub-band noise estimate satisfying a defined rule, reducing the current sub-band noise estimate by the defined amount for the at least one sub-band of the audio signal segment.

5. The method according to claim 4 , wherein the current sub-band noise estimate satisfies the defined rule when the current sub-band noise estimate exceeds a predefined value.

6. The method according to claim 4 , wherein the current sub-band noise estimate satisfies the defined rule when an energy level of an audio signal segment is less than a threshold level higher than a long term minimum energy level (lt_min) determined over a plurality of preceding audio signal segments.

7. The method according to claim 6 , wherein an energy level of an audio signal segment is determined to be less than the threshold level higher than lt_min based on information derived from an input audio signal and not based on use of information from a sound activity detector.

8. The method according to claim 1 , wherein reducing the current sub-band noise estimate by the defined amount comprises:
determining a step size to reduce the current sub-band noise estimate; and
reducing the current sub-band noise estimate by the step size.

9. An apparatus comprising:
at least one processor configured to perform operations comprising:
when a current energy level of an audio signal segment is higher than a current sub-band noise estimate and no pause is in the audio signal segment, reducing the current sub-band noise estimate by reducing the current sub-band noise estimate by a defined amount for at least one sub-band of the audio signal segment.

10. The apparatus according to claim 9 , wherein reducing the current sub-band noise estimate when the current energy level of the audio signal segment is higher than the current sub-band noise estimate and no pause is in the audio signal segment comprises reducing the current sub-band noise estimate by the defined amount for the at least one sub-band of the audio signal segment responsive to the audio signal segment comprising music when the current energy level of the audio signal segment is higher than the current sub-band noise estimate, and no pause is in the audio signal segment.

11. The apparatus according to claim 9 , wherein no pause is considered to in the audio signal segment when one or both of the following is fulfilled:
a predefined number of consecutive preceding audio signal segments have been determined to comprise an active signal;
a dynamic of an audio signal comprising an audio signal segment does not exceed a signal dynamics threshold.

12. The apparatus according to claim 9 , wherein the at least one processor configured to perform further operations comprising:
responsive to the current sub-band noise estimate satisfying a defined rule, reducing the current sub-band noise estimate by the defined amount for the at least one sub-band of the audio signal segment.

13. The apparatus according to claim 12 , wherein the current sub-band noise estimate satisfies the defined rule when the current sub-band noise estimate exceeds a predefined value.

14. The apparatus according to claim 12 , wherein the at least one processor performs further operations comprising setting a flag to indicate that the energy level of the audio signal segment is close to a long term minimum energy level (lt_min).

15. The apparatus according to claim 12 , wherein the current sub-band noise estimate satisfies the defined rule when an energy level of an audio signal segment is less than a threshold level higher than a long term minimum energy level (lt_min) determined over a plurality of preceding audio signal segments, wherein an energy level of an audio signal segment is determined to be less than the threshold level higher than lt_min based on information derived from an input audio signal, and not based on use of information from a sound activity detector.

16. The apparatus according to claim 12 , wherein in reducing the current sub-band noise estimate responsive to the current sub-band noise estimate satisfying the defined rule, the at least one processor performs operations comprising:
determining a step size to reduce the current sub-band noise estimate; and
reducing the current sub-band noise estimate by the step size.

17. A computer program product comprising a non-transitory computer readable storage medium storing instructions which, when executed on at least one processor, cause the at least one processor to perform operations comprising:
when a current energy level of an audio signal segment is higher than a current sub-band noise estimate and no pause is detected in the audio signal segment, reducing the current sub-band noise estimate by reducing the current sub-band noise estimate by a defined amount for at least one sub-band of the audio signal segment.

18. The computer program product of claim 17 , wherein reducing the current sub-band noise estimate when the current energy level of the audio signal segment is higher than the current sub-band noise estimate and no pause is detected in the audio signal segment comprises reducing the current sub-band noise estimate by the defined amount for the at least one sub-band of the audio signal segment responsive to the audio signal segment comprising music when the current energy level of the audio signal segment is higher than the current sub-band noise estimate, and no pause is detected in the audio signal segment.

19. The computer program product of claim 17 , wherein no pause is considered to in the audio signal segment when one or both of the following is fulfilled:
a predefined number of consecutive preceding audio signal segments have been determined to comprise an active signal;
a dynamic of an audio signal comprising an audio signal segment does not exceed a signal dynamics threshold.

20. The computer program product of claim 17 , wherein the non-transitory computer readable storage medium stores further instructions which, when executed on the at least one processor, cause the at least one processor to perform fur

Assignees

• Ericsson Telefon Ab L M

Inventors

• Sehlstedt Martin

Classifications

• G10L25/78

  Detection of presence or absence of voice signals (switching of direction of transmission by voice frequency in two-way loud-speaking telephone systems H04M9/10) · CPC title

• H04M1/62

  Constructional arrangements · CPC title

• G10L25/06

  the extracted parameters being correlation coefficients · CPC title

• H04W8/22

  Processing or transfer of terminal data, e.g. status or physical capabilities · CPC title

• G10L25/21 Primary

  the extracted parameters being power information · CPC title