Suspending noise cancellation using keyword spotting

Patent metadata

Field	Value
Publication number	US-9398367-B1
Application number	US-201414341222-A
Country	US
Kind code	B1
Filing date	Jul 25, 2014
Priority date	Jul 25, 2014
Publication date	Jul 19, 2016
Grant date	Jul 19, 2016

Abstract

Aspects of the disclosure provide suspension of noise cancellation at a noise-cancelling device using keyword spotting. In one aspect, a predetermined word or phrase can be spotted within an utterance received at the noise-cancelling device, and in response, noise cancellation can be suspended or otherwise terminated. The predetermined word or phrase can be specific to an end-user that utilizes the noise-cancelling device and/or a person that interacts with the end-user. In another aspect, interaction between an operator and the noise-cancelling device can be monitored after noise cancellation is suspended, and based at least on such interaction, a model for keyword spotting can be refined. In certain aspects, noise cancellation at the noise-cancelling device can be suspended or otherwise terminated in response to receiving a suspension directive via an electronic non-audio signal, for example, from a peripheral device. In other aspects, the noise-cancelling device can resume noise cancellation in response to a control signal.

First claim

What is claimed is:

1. A method, comprising:
receiving a first external audio signal from a microphone of a noise-cancelling device;
determining that the first external audio signal does not comprise a keyword corresponding to suspension of noise cancellation;
generating a noise-cancellation signal for cancelling the first external audio signal;
causing a speaker of the noise-cancelling device to generate audio corresponding to the noise-cancellation signal;
receiving a second external audio signal from the microphone of the noise-cancelling device;
determining that the second external audio signal comprises the keyword corresponding to suspension of noise cancellation;
facilitating presentation of at least a portion of the second external audio signal at the noise-cancelling device by suspending noise cancellation for at least the portion of the second external audio signal;
determining that a defined interaction with the noise cancelling device occurred, the defined interaction corresponding to an event to resume noise cancellation; and
resuming the noise cancellation for at least a second portion of the second external audio signal.

2. The method of claim 1 , further comprising receiving an input audio signal; and
wherein causing a speaker of the noise-cancelling device to generate audio corresponding to the noise-cancellation signal comprises combining at least a portion of the input audio signal with the noise-cancellation signal, and
causing the speaker to generate audio corresponding to the combined signal.

3. The method of claim 1 , wherein the applying the keyword model comprises applying at least one of a hidden Markov model or a Gaussian mixture model.

4. The method of claim 1 , further comprising further facilitating the presentation of at least the portion of the second external audio signal at the noise-cancelling device by causing the speaker to generate audio corresponding to at least the portion of the second external audio signal.

5. At least one non-transitory computer-readable medium comprising instructions, that when executed, cause a device to perform operations comprising:
receiving an external audio signal;
performing noise cancellation using a first portion of the external audio signal;
determining, using a keyword model, that a second portion of the external audio signal comprises a predetermined keyword;
causing the device to suspend noise cancellation for a third portion of the external audio signal;
determining occurrence of an interaction with the device to resume noise cancellation; and
resuming noise cancellation for a fourth portion of the external audio signal.

6. The at least one non-transitory computer-readable medium of claim 5 , wherein performing noise cancellation comprises generating a noise-cancellation signal for cancelling the first portion of the external audio signal, and
causing a speaker of the device to generate audio corresponding to the noise-cancellation signal.

7. The at least one non-transitory computer-readable medium of claim 5 , wherein the third portion of the external audio signal is after the second portion of the external audio signal.

8. The at least one non-transitory computer-readable medium of claim 5 , wherein the determining the occurrence of the interaction comprises receiving an indication to restart noise cancellation, and
the operation further comprising updating the keyword model in response to the indication.

9. The at least one non-transitory computer-readable of claim 6 , the operations further comprising receiving an input audio signal; and
wherein the causing the speaker of the device to generate audio corresponding to the noise-cancellation signal comprises combining at least a portion of the input audio signal with the noise-cancellation signal, and
causing the speaker to generate audio corresponding to the combined signal.

10. The at least one non-transitory computer-readable medium of claim 5 , wherein the predetermined keyword is associated with a person.

11. The at least one non-transitory computer-readable medium of claim 10 , the operations further comprising recognizing the person by using at least one of the first portion of the external audio signal or the second portion of the external audio signal.

12. The at least one non-transitory computer-readable of claim 5 , the operations further comprising causing the device to generate audio corresponding to the third portion of the external audio signal.

13. A noise-cancelling device, comprising:
an audio input unit configured to receive an external audio signal;
a noise cancellation unit configured to perform noise cancellation using a first portion of the external audio signal;
a keyword spotting unit configured to determine, using a keyword model, that a second portion of the external audio signal comprises a predetermined keyword;
a control unit configured to suspend noise cancellation for a third portion of the external audio signal; and
a monitoring unit configured to determine occurrence of an interaction with the device to restart noise cancellation;
wherein the control unit is further configured to restart noise cancellation for a fourth portion of the external audio signal.

14. The noise-cancelling device of claim 13 , wherein the noise cancellation unit is further configured to generate a noise-cancellation signal to cancel the first portion of the external audio signal, and
to cause a speaker of the noise-cancelling device to generate audio corresponding to the noise-cancellation signal.

15. The noise-cancelling device of claim 13 , wherein the third portion of the external audio signal is after the second portion of the external audio signal, and wherein the fourth portion of the external audio signal is after the third portion of the external audio signal.

16. The noise-cancelling device of claim 13 , wherein to determine the occurrence of the interaction, the monitoring unit is further configured to receive an indication to restart noise cancellation.

17. The noise-cancelling device of claim 16 , wherein a remote device updates the keyword model in response to the indication.

18. The noise-cancelling device of claim 14 , wherein the audio input unit is further configured to receive an input audio signal; and
the noise cancellation unit is further configured to combine at least a portion of the input audio signal with the noise-cancellation signal, and to cause the speaker to generate audio corresponding to the combined signal.

19. The noise-cancelling device of claim 13 , wherein the predetermined keyword is associated with a person.

20. The noise-cancelling device of claim 19 , wherein the keyword spotting unit is further configured to identify the person by performing a spectral analysis of at least one of the first portion of the external audio signal or the second portion of the external audio signal.

21. The noise-cancelling device of claim 13 , wherein the noise-cancelling device comprises one of a noise-cancelling headphone, a noise-cancelling earbud, or a portable noise-cancelling apparatus.

Assignees

• Amazon Tech Inc

Inventors

• Scott Benjamin David
• Rafn Mark Edward

Classifications

• H04R1/1083 Primary

  Reduction of ambient noise (active noise reduction per se G10K11/175; protective devices for the ear, e.g. providing acoustic protection A61F11/06) · CPC title

• H04R2460/01

  Hearing devices using active noise cancellation · CPC title

• G10L2015/088

  Word spotting · CPC title

• G10L17/00

  Speaker identification or verification techniques · CPC title