SPEAKER RECOGNITION

• US 20170092278A1
• Filed: 05/24/2016
• Published: 03/30/2017
• Est. Priority Date: 09/30/2015
• Status: Abandoned Application

• Current Assignee: Apple Inc.
• Original Assignee: Apple Inc.
• Inventors: MCALLASTER, Donald R., EVERMANN, Gunnar
• Application Number:
  US15/163,392
• Publication Number:
  US 20170092278A1
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 17/04 Training, enrolment or model building
• CPC Class Codes: G10L 17/08 Use of distortion metrics or a particular distance between probe pattern and reference templates
• CPC Class Codes: G10L 17/24 the user being prompted to utter a password or a predefined phrase

Abstract

A non-transitory computer-readable storage medium stores one or more programs including instructions, which when executed by an electronic device, cause the electronic device to receive natural-language speech input from one of a plurality of users, the natural-language speech input having a set of acoustic properties; and determine whether the natural-language speech input corresponds to both a user-customizable lexical trigger and a set of acoustic properties associated with the user; where in accordance with a determination that the natural language speech input corresponds to both a user-customizable lexical trigger and a set of acoustic properties associated with the user, invoke a virtual assistant; and in accordance with a determination that either the natural language speech input fails to correspond to a user-customizable lexical trigger or the natural-language speech input fails to have a set of acoustic properties associated with the user, forego invocation of a virtual assistant.

Claims

1. A non-transitory computer-readable storage medium storing one or more programs, the one or more programs comprising instructions, which when executed by an electronic device, cause the electronic device to:

receive natural-language speech input from one of a plurality of users, the natural-language speech input having a set of acoustic properties; and

determine whether the natural-language speech input corresponds to both a user-customizable lexical trigger and a set of acoustic properties associated with the user;

wherein

in accordance with a determination that the natural language speech input corresponds to both a user-customizable lexical trigger and a set of acoustic properties associated with the user, invoke a virtual assistant; and

in accordance with a determination that either the natural language speech input fails to correspond to a user-customizable lexical trigger or the natural-language speech input fails to have a set of acoustic properties associated with the user, forego invocation of a virtual assistant.

2. The non-transitory computer-readable storage medium storing one or more programs, the one or more programs comprising instructions, of claim 1, the one or more programs further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

receive user input of at least one word; and

store the at least one word as the lexical trigger.

3. The non-transitory computer-readable storage medium storing one or more programs, the one or more programs comprising instructions, of claim 1, the one or more programs further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

further in accordance with a determination that the natural language speech input corresponds to both a user-customizable lexical trigger and a set of acoustic properties associated with the user;

compare the acoustic properties of the received natural-language speech input of the user to a reference set of acoustic properties accessible to the virtual assistant; and

store the differences between the acoustic properties of the received natural-language speech input of the user and the reference set of acoustic properties.

4. The non-transitory computer-readable storage medium storing one or more programs, the one or more programs comprising instructions, of claim 1, the one or more programs further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

further in accordance with a determination that the natural language speech input corresponds to both a user-customizable lexical trigger and a set of acoustic properties associated with the user;

request the user to speak at least one preselected word;

in response to the request, receive natural-language speech input of the user speaking the one or more preselected words.

5. The non-transitory computer-readable storage medium storing one or more programs, the one or more programs comprising instructions, of claim 1, the instructions to determine whether the natural-language speech input corresponds to both a user-customizable lexical trigger and a set of acoustic properties associated with the user further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

determine whether the set of acoustic properties of the natural-language speech input matches the set of acoustic properties of one of the plurality of user profiles accessible to the virtual assistant;

in accordance with a determination that the set of acoustic properties of the natural-language speech input matches the set of acoustic properties of one of the plurality of user profiles, conclude that the natural language speech input corresponds to a set of acoustic properties associated with the user; and

in accordance with a determination that the input fails to match any of the plurality of user profiles, proceed to forego invocation of the virtual assistant.

6. The non-transitory computer-readable storage medium storing one or more programs, the one or more programs comprising instructions, of claim 5, the one or more programs further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

create a user profile for at least one of a plurality of users of the electronic device, the user profile including a user identity; and

store the at least one user profile.

7. The non-transitory computer-readable storage medium storing one or more programs, the one or more programs comprising instructions, of claim 5, the one or more programs further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

receive a user profile for at least one of a plurality of users of the electronic device, the user profile including a user identity.

8. The non-transitory computer-readable storage medium of claim 5, the non-transitory computer-readable storage medium further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

first determine whether the natural-language speech input matches a set of acoustic properties associated with at least one of the plurality of user profiles; and

in accordance with a determination that the natural-language speech input matches a set of acoustic properties associated with one of the plurality of user profiles, proceed to determine whether the natural language speech input matches the user-customizable lexical trigger; and

in accordance with a determination that the natural-language speech input fails to match any of the plurality of user profiles, proceed to forego invocation of the virtual assistant.

9. The non-transitory computer-readable storage medium of claim 5, the non-transitory computer-readable storage medium further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

receive natural-language speech input of the user other than repetition of preselected words;

compare the acoustic properties of the received natural-language speech input of the user to the acoustic properties of the received natural-language speech input stored in the user profile; and

determine whether the acoustic properties of the received natural-language speech input of the user substantially differ from the acoustic properties of the received natural-language speech input stored in the user profile;

in accordance with a determination that the acoustic properties of the received natural-language speech input of the user substantially differ from the acoustic properties of the received natural-language speech input stored in the user profile;

update the user profile of the user based on the acoustic properties of the received natural-language speech input of the user; and

store the updated user profile; and

in accordance with a determination that the acoustic properties of the received natural-language speech input of the user do not substantially differ from the acoustic properties of the received natural-language speech input stored in the user profile, forego the update of the user profile based on the acoustic properties of the received natural-language speech input of the user.

10. The non-transitory computer-readable storage medium storing one or more programs, the one or more programs comprising instructions, of claim 1, the one or more programs further comprising instructions which when executed by the one or more processors of the electronic device, cause the device to:

transmit at least one user profile from the electronic device.

11. The non-transitory computer-readable storage medium of claim 1, the non-transitory computer-readable storage medium further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

further in accordance with a determination that the natural-language speech input corresponds to both a user-customizable lexical trigger and a set of acoustic properties associated with the user, receive at least one additional security identifier; and

determine whether the at least one additional security identifier is associated with the user;

in accordance with a determination that the at least one additional security identifier is associated with the user, invoke the virtual assistant;

in accordance with a determination that the at least one additional security identifier is not associated with the user, forego invocation of the virtual assistant.

12. The non-transitory computer-readable storage medium of claim 1, the non-transitory computer-readable storage medium further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

enroll at least one user;

wherein the instructions to enroll at least one user further comprise instructions, which when executed by the one or more processors of the electronic device, cause the device to;

request the user to say one or more preselected words;

in response to the request, receive user input that includes natural-language speech input corresponding to the one or more preselected words.

13. The non-transitory computer-readable storage medium of claim 1, the non-transitory computer-readable storage medium further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

enroll at least one user during the first use of the electronic device by the user.

14. The non-transitory computer-readable storage medium of claim 1, the non-transitory computer-readable storage medium further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

update enrollment of at least one user upon a detected change to the acoustic properties of the voice of the user.

15. The non-transitory computer-readable storage medium of claim 14, the non-transitory computer-readable storage medium further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

request at least one additional security identifier from the user to perform the enrollment; and

determine if the at least one additional security identifier is associated with the user;

in accordance with a determination that the at least one additional security identifier is associated with the user, enroll the user;

in accordance with a determination that the at least one additional security identifier is not associated with the user, forego enrollment of the user.

16. The non-transitory computer-readable storage medium of claim 1, the non-transitory computer-readable storage medium further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

receive natural-language speech input that corresponds to a set of acoustic properties associated with the user but not the user-customizable lexical trigger;

in response to the receipt of natural-language speech input that corresponds to one but not both of a set of acoustic properties associated with the user and the user-customizable lexical trigger, request the user to repeat the natural-language speech input; and

determine whether the repeated natural-language speech input corresponds to both a user-customizable lexical trigger and a set of acoustic properties associated with the user;

wherein

in accordance with a determination that the natural language speech input corresponds to both a user-customizable lexical trigger and a set of acoustic properties associated with the user;

invoke a virtual assistant; and

enroll the first natural-language speech input of the user; and

in accordance with a determination that either the natural-language speech input fails to correspond to a user-customizable lexical trigger or the natural-language speech input fails to have a set of acoustic properties associated with the user, forego invocation of a virtual assistant.

17. The non-transitory computer-readable storage medium of claim 1, the instructions to determine whether the natural-language speech input corresponds to both a user-customizable lexical trigger and a set of acoustic properties associated with the user further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

store one or more supervectors, each associated with the acoustic properties of the voice of a user;

generate a supervector based on the natural-language speech input;

compare the generated supervector with one or more stored supervectors to generate a score; and

determine whether the score exceeds a threshold;

in accordance with a determination that the score exceeds the threshold, conclude that the natural-language speech input corresponds to a set of acoustic properties associated with a user; and

in accordance with a determination that the score fails to exceed the threshold, conclude that the natural-language speech input fails to correspond to a set of acoustic properties associated with a user.

18. The non-transitory computer-readable storage medium of claim 16, the instructions to generate a supervector further comprising instructions, which when executed by the one or more processors of the electronic device, cause the device to:

generate the supervector by utilizing state backtrace.

19. A method of using a virtual assistant, comprising:

at an electronic device configured to transmit and receive data,

receiving natural-language speech input from one of a plurality of users, the natural-language speech input having a set of acoustic properties; and

determining whether the natural-language speech input corresponds to both a user-customizable lexical trigger and a set of acoustic properties associated with the user;

wherein

in accordance with a determination that the natural language speech input corresponds to both a user-customizable lexical trigger and a set of acoustic properties associated with the user, invoking a virtual assistant; and

in accordance with a determination that either the natural language speech input fails to correspond to a user-customizable lexical trigger or the natural-language speech input fails to have a set of acoustic properties associated with the user, foregoing invocation of a virtual assistant.

20. An electronic device, comprising:

a memory;

a microphone; and

a processor coupled to the memory and the microphone, the processor configured to;

receive natural-language speech input from one of a plurality of users, the natural-language speech input having a set of acoustic properties; and

determine whether the natural-language speech input corresponds to both a user-customizable lexical trigger and a set of acoustic properties associated with the user;

wherein

in accordance with a determination that the natural language speech input corresponds to both a user-customizable lexical trigger and a set of acoustic properties associated with the user, invoke a virtual assistant; and

in accordance with a determination that either the natural language speech input fails to correspond to a user-customizable lexical trigger or the natural-language speech input fails to have a set of acoustic properties associated with the user, forego invocation of a virtual assistant.