Wakeword detection using a secondary microphone

Patent metadata

Field	Value
Publication number	US-10997971-B2
Application number	US-201916272898-A
Country	US
Kind code	B2
Filing date	Feb 11, 2019
Priority date	Feb 11, 2019
Publication date	May 4, 2021
Grant date	May 4, 2021

Abstract

Techniques for capturing spoken user inputs while a device is prevented from capturing such spoken user inputs are described. When a first device has a status representing it is unbeneficial for the first device to perform wakeword detection, a second device (e.g. a vehicle) may perform wakeword detection on behalf of the first device. The second device may be unable to send audio data, representing a spoken user input, to a speech processing system. In such an example, the second device may send the audio data to a third device, which may send the audio data to the speech processing system.

First claim

What is claimed is:

1. A system, comprising:
at least one processor; and
at least one memory comprising instructions that, when executed by the at least one processor, cause the system to:
receive a first indicator representing a device status of a first device corresponding to a first device identifier;
receive a vehicle identifier corresponding to a vehicle to which the first device is connected;
send, to the first device, a first instruction to cease processing first audio to detect a spoken wakeword;
store an association between the first device identifier and the vehicle identifier;
receive audio data from a second device, the audio data representing a spoken user input, the second device being connected to the vehicle using a Bluetooth component;
after sending the first instruction, receive the vehicle identifier from the second device;
determine the vehicle identifier is associated with the first device identifier; and
perform speech processing on the audio data using a first speech processing system component associated with the first device instead of a second speech processing system component associated with the second device.

2. The system of claim 1 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to:
receive a second indicator representing a second device status of the first device;
after receiving the second indicator, receive, from the first device, a third device identifier corresponding to a third device connected to the first device using a second Bluetooth component;
determine a profile associated with a plurality of device identifiers including the first device identifier;
determine the profile is associated with the third device identifier; and
send, to the first device, a second instruction to cease processing of second audio to detect the spoken wakeword.

3. The system of claim 1 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to:
determine a wakeword detection model associated with the first device identifier, the wakeword detection model being generated using training audio data associated with a user identifier, the user identifier being associated with the first device identifier; and
send, to the vehicle, the wakeword detection model.

4. The system of claim 3 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to:
receive a second indicator representing a second device status of the first device;
determine the second device status represents the first device is to be permitted to process third audio to detect the spoken wakeword; and
send, to the vehicle, a third instruction to disable the wakeword detection model.

5. A system, comprising:
at least one processor; and
at least one memory comprising instructions that, when executed by the at least one processor, cause the system to:
receive an indicator representing a device status of a first device corresponding to a first device identifier;
determine a second device configured to perform wakeword detection;
receive audio data from a third device associated with the second device, the audio data representing a spoken user input; and
perform speech processing on the audio data as if the audio data was received from the first device.

6. The system of claim 5 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to:
based at least in part on determining the second device, send, to the first device, an instruction to cease performing wakeword detection.

7. The system of claim 5 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to:
based at least in part on receiving the indicator, send, to the first device, an instruction to cause the first device to cease performing wakeword detection.

8. The system of claim 5 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to:
receive a second indicator representing a second device status of the first device;
based at least in part on the second device status, receive, from the first device, a fourth device identifier corresponding to a fourth device connected to the first device via a wireless connection; and
cause the first device to cease performing wakeword detection.

9. The system of claim 5 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to:
receive a second indicator representing a second device status of the first device;
based at least in part on the second device status, determine a profile associated with a plurality of device identifiers including the first device identifier and a fourth device identifier corresponding to a fourth device;
determine, in the profile, that the first device identifier is associated with first position data;
determine, in the profile, that the fourth device identifier is associated with second position data substantially corresponding to the first position data; and
cause the first device to cease performing wakeword detection.

10. The system of claim 5 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to:
receive a second indicator representing a second device status of the first device;
based at least in part on the second device status, determine a profile associated with a plurality of device identifiers including the first device identifier and a fourth device identifier corresponding to a fourth device;
determine, in the profile, that the first device identifier is associated with network identifying information;
determine, in the profile, that the fourth device identifier is associated with the network identifying information; and
cause the first device to cease performing wakeword detection.

11. The system of claim 5 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to:
determine a wakeword detection model associated with the first device identifier; and
send, to the second device, the wakeword detection model.

12. The system of claim 11 , wherein the at least one memory further comprises instructions that, when executed by the at least one processor, further cause the system to:
receive a second indicator representing a second device status of the first device; and
based at least in part on the second indicator, send, to the second device, a second instruction to disable the wakeword detection model.

13. A method, comprising:
receiving an indicator representing a device status of a first device corresponding to a first device identifier;
determining a second device configured to perform wakeword detection;
receiving audio data from a third device associated with the second device, the audio data representing a spoken user input; and
performing speech processing on the audio data as if the audio data was received from the first device.

14. The method of claim 13 , further comprising:
based at least in part on determining the second device, sending, to the first device, an instruction to cease performing wakeword detection.

15. The method of claim 13 , further comprising:
b

Assignees

• Amazon Tech Inc

Inventors

• Mitchell Andrew
• Nagy Gabor

Classifications

• G10L15/22 Primary

  Procedures used during a speech recognition process, e.g. man-machine dialogue · CPC title

• G10L2015/088

  Word spotting · CPC title

• G10L2015/223

  Execution procedure of a spoken command · CPC title

• G10L15/08 Primary

  Speech classification or search · CPC title

• G10L2015/228

  of application context · CPC title