User-specific noise suppression for voice quality improvements

• US 10,446,167 B2
• Filed: 01/27/2014
• Issued: 10/15/2019
• Est. Priority Date: 06/04/2010
• Status: Active Grant

• Current Assignee: Apple Inc.
• Original Assignee: Apple Inc.
• Inventors: Lindahl, Aram, Paquier, Baptiste Pierre
• Primary Examiner(s): Opsasnick, Michael N
• Application Number:
  US14/165,523
• Publication Number:
  US 20140142935A1
• Time in Patent Office: 2,087 Days
• Field of Search: 704275
• CPC Class Codes: G10L 21/0208 Noise filtering

Abstract

Systems, methods, and devices for user-specific noise suppression are provided. For example, when a voice-related feature of an electronic device is in use, the electronic device may receive an audio signal that includes a user voice. Since noise, such as ambient sounds, also may be received by the electronic device at this time, the electronic device may suppress such noise in the audio signal. In particular, the electronic device may suppress the noise in the audio signal while substantially preserving the user voice via user-specific noise suppression parameters. These user-specific noise suppression parameters may be based at least in part on a user noise suppression preference or a user voice profile, or a combination thereof.

Claims

1. A method, comprising:

at an electronic device with one or more processors and memory;

receiving an audio signal that includes a voice of a near-end user of the electronic device when a voice-related feature of the electronic device is in use;

suppressing noise in the audio signal using the electronic device while substantially preserving the voice of the near-end user based at least in part on user-specific noise suppression parameters,

wherein the user-specific noise suppression parameters are based at least in part on a user noise suppression preference,

wherein the user noise suppression preference is based on a user selection between a first filtered audio signal generated by applying a first set of noise suppression parameters to a test audio signal and a second filtered audio signal generated by applying a second set of noise suppression parameters to the test audio signal, and

wherein the test audio signal includes speech of the near-end user and one or more distracters, the test audio signal being output by the electronic device responsive to receiving a user input indicating initiation of a voice training mode of the electronic device; and

transmitting the audio signal to a remote device for receipt by a far-end user.

2. The method of claim 1, wherein the user-specific noise suppression parameters suppress noise in the audio signal while substantially preserving the voice at least in part by amplifying frequencies associated with a user voice profile.

3. The method of claim 1, wherein the user-specific noise suppression parameters suppress noise in the audio signal while substantially preserving the voice at least in part by suppressing frequencies not associated with a user voice profile.

4. The method of claim 1, wherein a first noise suppression strength associated with the first set of noise suppression parameters is greater than a second noise suppression strength associated with the second set of noise suppression parameters.

5. The method of claim 1, further comprising:

determining a user voice profile based on the audio signal; and

determining the user-specific noise suppression parameters based on the user voice profile.

6. The method of claim 1, further comprising:

adjusting the user-specific noise suppression parameters based on user input received while receiving the audio signal.

7. The method of claim 1, wherein the first filtered audio signal is generated after applying a third set of noise suppression parameters to the test audio signal and wherein the second filtered audio signal is generated after applying a fourth set of noise suppression parameters to the test audio signal.

8. The method of claim 1, wherein the user selection includes a selection of a displayed menu item.

9. The method of claim 1, wherein the user selection is responsive to a prompt, output by the electronic device, prompting user selection between the first filtered audio signal and the second filtered audio signal.

10. The method of claim 1, wherein the user noise suppression preference is based at least in part on a user noise suppression training sequence.

11. The method of claim 10, wherein the user noise suppression training sequence comprises receiving at the electronic device a user selection of preferred noise parameters after noise suppression parameters have been tested on a second test audio signal and played back to the near-end user.

12. The method of claim 10, wherein the user noise suppression training sequence comprises testing noise suppression parameters as applied to a test audio signal that includes a voice sample of the near-end user and at least one distractor.

13. The method of claim 1, wherein the user noise suppression preference is further based on a user-selected noise suppression setting.

14. The method of claim 13, wherein the user-selected noise suppression setting comprises a noise suppression strength setting.

15. The method of claim 13, wherein the user-selected noise suppression setting is user selectable in real time while the voice-related feature of the electronic device is in use.

16. An electronic device, comprising:

one or more processors;

one or more microphones; and

memory storing one or more programs for execution by the at least one processor, the one or more programs including instructions for;

receiving an audio signal that includes a voice of a near-end user of the electronic device when a voice-related feature of the electronic device is in use;

suppressing noise in the audio signal using the electronic device while substantially preserving the voice of the near-end user based at least in part on user-specific noise suppression parameters,

wherein the user-specific noise suppression parameters are based at least in part on a user noise suppression preference,

wherein the user noise suppression preference is based on a user selection between a first filtered audio signal generated by applying a first set of noise suppression parameters to a test audio signal and a second filtered audio signal generated by applying a second set of noise suppression parameters to the test audio signal, and

wherein the test audio signal includes speech of the near-end user and one or more distractors, the test audio signal being output by the electronic device responsive to receiving a user input indicating initiation of a voice training mode of the electronic device; and

transmitting the audio signal to a remote device for receipt by a far-end user.

17. The electronic device of claim 16, wherein the user noise suppression preference is further based on a user-selected noise suppression setting, and

wherein the user-selected noise suppression setting is user selectable in real time while the voice-related feature of the electronic device is in use.

18. The electronic device of claim 16, wherein the user-specific noise suppression parameters suppress noise in the audio signal while substantially preserving the voice at least in part by amplifying frequencies associated with a user voice profile.

19. The electronic device of claim 16, wherein the user-specific noise suppression parameters suppress noise in the audio signal while substantially preserving the voice at least in part by suppressing frequencies not associated with a user voice profile.

20. The electronic device of claim 16, wherein the one or more programs further include instructions for:

determining a user voice profile based on the audio signal; and

determining the user-specific noise suppression parameters based on the user voice profile.

21. The electronic device of claim 16, wherein the user selection is responsive to a prompt, output by the electronic device, prompting user selection between the first filtered audio signal and the second filtered audio signal.

22. A non-transitory computer-readable storage medium, storing one or more programs for execution by one or more processors of an electronic device, the one or more programs including instructions for:

receiving an audio signal that includes a voice of a near-end user of the electronic device when a voice-related feature of the electronic device is in use;

suppressing noise in the audio signal using the electronic device while substantially preserving the voice of the near-end user based at least in part on user-specific noise suppression parameters,

wherein the user-specific noise suppression parameters are based at least in part on a user noise suppression preference,

wherein the user noise suppression preference is based on a user selection between a first filtered audio signal generated by applying a first set of noise suppression parameters to a test audio signal and a second filtered audio signal generated by applying a second set of noise suppression parameters to the test audio signal, and

wherein the test audio signal includes speech of the near-end user and one or more distracters, the test audio signal being output by the electronic device responsive to receiving a user input indicating initiation of a voice training mode of the electronic device; and

transmitting the audio signal to a remote device for receipt by a far-end user.

23. The non-transitory computer-readable storage medium of claim 22, wherein the user noise suppression preference is further based on a user-selected noise suppression setting, and

wherein the user-selected noise suppression setting is user selectable in real time while the voice-related feature of the electronic device is in use.

24. The non-transitory computer-readable storage medium of claim 22, wherein the user-specific noise suppression parameters suppress noise in the audio signal while substantially preserving the voice at least in part by amplifying frequencies associated with a user voice profile.

25. The non-transitory computer-readable storage medium of claim 22, wherein the user-specific noise suppression parameters suppress noise in the audio signal while substantially preserving the voice at least in part by suppressing frequencies not associated with a user voice profile.

26. The non-transitory computer-readable storage medium of claim 22, wherein the one or more programs further include instructions for:

determining a user voice profile based on the audio signal; and

determining the user-specific noise suppression parameters based on the user voice profile.

27. The non-transitory computer-readable storage medium of claim 22, wherein the user selection is responsive to a prompt, output by the electronic device, prompting user selection between the first filtered audio signal and the second filtered audio signal.

28. An electronic device, comprising:

one or more processors;

one or more microphones; and

memory storing one or more programs for execution by the one or more processors, the one or more programs including instructions for;

determining whether a signal-to-noise ratio of a first audio signal obtained while a voice-related feature of the electronic device is in use exceeds a threshold;

in accordance with a determination that the signal-to-noise ratio exceeds the threshold, obtaining a user voice sample from the first audio signal;

determining a user voice profile based at least in part on the user voice sample;

determining user-specific noise suppression parameters based at least in part on the user voice profile;

obtaining a second audio signal that includes a user voice and ambient sounds; and

applying noise suppression to the second audio signal based at least in part on the user-specific noise suppression parameters to suppress the ambient sounds of the second audio signal.

29. The electronic device of claim 28, wherein the one or more programs further include instructions for:

determining the user voice profile further based at least in part on another user voice sample obtained during an activation period of the electronic device.

30. The electronic device of claim 28, wherein the one or more programs further include instructions for:

determining whether the user voice corresponds to a known user and, when the user voice corresponds to the known user, recalling the user voice profile associated with the user voice.

31. The electronic device of claim 28, wherein the one or more programs further include instructions for:

determining whether the user voice corresponds to a known user, wherein determining whether the signal-to-noise ratio of the first audio signal obtained while the voice-related feature of the electronic device is in use exceeds the threshold is performed in accordance with a determination that the user voice does not correspond to the known user.

32. A method, comprising:

at an electronic device with one or more processors and memory;

determining whether a signal-to-noise ratio of a first audio signal obtained while a voice-related feature of the electronic device is in use exceeds a threshold;

in accordance with a determination that the signal-to-noise ratio exceeds the threshold, obtaining a user voice sample from the first audio signal;

determining a user voice profile based at least in part on the user voice sample;

determining user-specific noise suppression parameters based at least in part on the user voice profile;

obtaining a second audio signal that includes a user voice and ambient sounds; and

applying noise suppression to the second audio signal based at least in part on the user-specific noise suppression parameters to suppress the ambient sounds of the second audio signal.

33. The method of claim 32, further comprising:

determining the user voice profile further based at least in part on another user voice sample obtained during an activation period of the electronic device.

34. The method of claim 32, further comprising:

determining whether the user voice corresponds to a known user and, when the user voice corresponds to the known user, recalling the user voice profile associated with the user voice.

35. The method of claim 32, further comprising:

determining whether the user voice corresponds to a known user, wherein determining whether the signal-to-noise ratio of the first audio signal obtained while the voice-related feature of the electronic device is in use exceeds the threshold is performed in accordance with a. determination that the user voice does not correspond to the known user.

36. A non-transitory computer-readable storage medium, storing one or more programs for execution by one or more processors of an electronic device, the one or more programs including instructions for:

determining whether a signal-to-noise ratio of a first audio signal obtained while a voice-related feature of the electronic device is in use exceeds a threshold;

in accordance with a determination that the signal-to-noise ratio exceeds the threshold, obtaining a user voice sample from the first audio signal;

determining a user voice profile based at least in part on the user voice sample;

determining user-specific noise suppression parameters based at least in part on the user voice profile;

obtaining a second audio signal that includes a user voice and ambient sounds; and

applying noise suppression to the second audio signal based at least in part on the user-specific noise suppression parameters to suppress the ambient sounds of the second audio signal.

37. The non-transitory computer-readable storage medium of claim 36, wherein the one or more programs further include instructions for:

determining the user voice profile further based at least in part on another user voice sample obtained during an activation period of the electronic device.

38. The non-transitory computer-readable storage medium of claim 36, wherein the one or more programs further include instructions for:

determining whether the user voice corresponds to a known user and, when the user voice corresponds to the known user, recalling the user voice profile associated with the user voice.

39. The non-transitory computer-readable storage medium of claim 36, wherein the one or more programs further include instructions for:

determining whether the user voice corresponds to a known user, wherein determining whether the signal-to-noise ratio of the first audio signal obtained while the voice-related feature of the electronic device is in use exceeds the threshold is performed in accordance with a determination that the user voice does not correspond to the known user.