Method And Apparatus For Voice Activity Determination

• US 20120310641A1
• Filed: 08/13/2012
• Published: 12/06/2012
• Est. Priority Date: 04/25/2008
• Status: Active Grant

• Current Assignee: Nokia Corporation
• Original Assignee: Nokia Corporation
• Inventors: Niemistö, Riitta Elina, Valve, Päivi Marianna
• Granted Patent: US 8,682,662 B2
• Time in Patent Office: Days
• US Class Current: 704/233
• 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 2021/02166 Microphone arrays; Beamforming
• CPC Class Codes: G10L 25/78 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

Abstract

In accordance with an example embodiment of the invention, there is provided an apparatus for detecting voice activity in an audio signal. The apparatus comprises a first voice activity detector for making a first voice activity detection decision based at least in part on the voice activity of a first audio signal received from a first microphone. The apparatus also comprises a second voice activity detector for making a second voice activity detection decision based at least in part on an estimate of a direction of the first audio signal and an estimate of a direction of a second audio signal received from a second microphone. The apparatus further comprises a classifier for making a third voice activity detection decision based at least in part on the first and second voice activity detection decisions.

Claims

1-13. -13. (canceled)

14. An apparatus, comprising:

at least two microphones configured to receive a first acoustic signal and a second acoustic signal;

at least one processor; and

at least one memory including computer program code

the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to perform at least the following;

produce a first beam signal and a second beam signal determined from the first and second acoustic signals; and

determine a voice activity detection decision based at least in part on a characteristic of the first and second acoustic signals and/or a relationship of the first and second beam signals.

15. The apparatus of claim 14 wherein the characteristic of the first and second acoustic signals comprises a directional characteristic of the first and second acoustic signals.

16. The apparatus of claim 14 wherein the relationship of the first and second beam signals comprises a ratio of the first and second beam signals.

17. The apparatus of claim 14 wherein the determined voice activity detection decision is based at least in part on the characteristic of the first and second acoustic signals and the relationship of the first and second beam signals, and wherein the first and second beam signals comprise main beam and anti beam signals.

18. The apparatus of claim 14 wherein the determined voice activity detection decision is based on an estimated direction of the first or second acoustic signal, and on a ratio of the first and second beam signals.

19. The apparatus of claim 14 wherein the determined voice activity detection decision is based on:

a quotient of differences comprising a difference between a signal power of the first acoustic signal and a signal power of the first beam signal and a difference between a signal power of the second acoustic signal and a signal power of the second beam signal; and

a ratio of the first and second beam signals.

20. The apparatus of claim 14 wherein a noise cancellation circuit may estimate and update a background noise spectrum when a voice activity decision indication indicates that the audio signal does not contain speech.

21. The apparatus of claim 14 wherein the first beam signal is a main beam signal, and wherein the second beam signal is an anti beam signal.

22. The apparatus of claim 14 wherein the apparatus is a mobile phone.

23. An apparatus comprising:

at least one first microphone configured to receive a first acoustic signal;

at least one second microphone configured to receive a second acoustic signal;

a beam former configured to produce a main beam signal and an anti beam signal calculated from the first and second acoustic signals; and

a voice activity detector connected to the at least one first microphone and the at least one second microphone, wherein the voice activity detector is configured to make a voice activity detection decision based at least in part on a characteristic of the first and second acoustic signals and/or a relationship of the main beam and anti beam signals.

24. The apparatus of claim 23 wherein the characteristic of the first and second acoustic signals comprises a directional characteristic of the first and second acoustic signals.

25. The apparatus of claim 23 wherein the relationship of the main beam and anti beam signals comprises a ratio of the main beam and anti beam signals.

26. The apparatus of claim 23 wherein the voice activity detector is configured to make a voice activity detection decision based on the characteristic of the first and second acoustic signals and the relationship of the main beam and anti beam signals, and wherein the characteristic of the first and second acoustic signals comprises a directional characteristic of the first and second acoustic signals.

27. The apparatus of claim 23 wherein the voice activity detector is configured to make the voice activity detection decision based at least in part on signal powers of the first acoustic signal and the second acoustic signal and signal powers of the main beam signal and the anti beam signal.

28. The apparatus of claim 23 further comprising a classifier, wherein the classifier is configured to make another voice activity detection decision based at least in part on the determined voice activity detection decision.

29. The apparatus of claim 28 wherein the classifier is configured to classify the acoustic signals as speech if the voice activity detector detects voice activity in the acoustic signals, and wherein the classifier is configured to classify the acoustic signals as non-speech if the voice activity detector detects non-speech activity for a predetermined duration of time.

30. A method comprising:

receiving a first acoustic signal at a first microphone;

receiving a second acoustic signal at a second microphone;

producing a first beam signal and a second beam signal determined from the first and second acoustic signals; and

determining a voice activity detection decision based at least in part on a characteristic of the first and second acoustic signals and/or a relationship of the first and second beam signals.

31. The method of claim 30 wherein the characteristic of the first and second acoustic signals comprises a directional characteristic of the first and second acoustic signals.

32. The method of claim 30 wherein the relationship of the first and second beam signals comprises a ratio of main beam and anti beam signals.

33. The method of claim 30 wherein the determined voice activity detection decision is based at least in part on the characteristic of the first and second acoustic signals and the relationship of the first and second beam signals, and wherein the relationship of the first and second beam signals comprises a ratio of the first and second beam signals.