Background speech recognition assistant using speaker verification

• US 9,142,219 B2
• Filed: 05/16/2014
• Issued: 09/22/2015
• Est. Priority Date: 09/27/2011
• Status: Active Grant

• Current Assignee: Sensory Incorporated
• Original Assignee: Sensory Incorporated
• Inventors: Mozer, Todd F.
• Primary Examiner(s): Godbold, Douglas
• Application Number:
  US14/280,261
• Publication Number:
  US 20140257812A1
• Time in Patent Office: 494 Days
• Field of Search: 704246-250, 704/270, 704/270.1
• 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 17/00 Speaker identification or verification techniques
• CPC Class Codes: G10L 17/22 Interactive procedures; Man-machine interfaces
• CPC Class Codes: G10L 2015/227 of the speaker; Human-factor methodology

Abstract

In one embodiment, a method includes receiving an acoustic input signal at a speech recognizer. A user is identified that is speaking based on the acoustic input signal. The method then determines speaker-specific information previously stored for the user and a set of responses based on the recognized acoustic input signal and the speaker-specific information for the user. It is determined if the response should be output and the response is outputted if it is determined the response should be output.

Claims

1. A method comprising:

receiving, by a computing device, an acoustic input signal at a speech recognizer;

identifying, by the computing device, a user that is speaking based on the acoustic input signal;

recognizing, by the computing device via the speech recognizer, speech uttered by the user in the acoustic input signal;

determining, by the computing device, speaker-specific information previously stored for the user;

determining, by the computing device, a set of potential responses based on the recognized speech and the speaker-specific information for the user;

ranking, by the computing device, the set of potential responses based on one or more criteria and the speaker-specific information;

determining, by the computing device for each response in the set of potential responses, whether the response should be output or should not be output based on the response'"'"'s ranking; and

if the response should be output;

selecting, by the computing device from among a plurality of preconfigured output methods, an output method for outputting the response to the user, the selecting being based on the response'"'"'s ranking; and

outputting, by the computing device, the response to the user using the selected output method.

2. The method of claim 1 wherein the speech recognizer is configured to recognize the acoustic input signal in an always on mode, and wherein the response is outputted without touching the computing device or speaking a trigger phrase to activate the speech recognizer.

3. The method of claim 1 wherein the response is outputted after a user speaks a trigger phrase to activate the speech recognizer.

4. The method of claim 1 wherein the speech recognizer operates in an always on mode, and wherein the speech recognizer identifies the user upon receiving a trigger phrase.

5. The method of claim 1 wherein the speaker-specific information is associated with previous speech recognition of speech from the user.

6. The method of claim 1 further comprising:

determining a set of classifications based on the speaker-specific information;

classifying portions of the acoustic input signal into different classifications in the set of classifications;

selecting a classification in the set of classifications based on a criterion associated with the classification; and

using the classification to determine the set of potential responses.

7. The method of claim 6 wherein the speaker-specific information is used to modify a classification in the set of classifications based on a preference of the user in the speaker-specific information.

8. The method of claim 7 wherein a set of keywords associated with the speaker-specific information is used in the classification.

9. The method of claim 6 wherein classifying portions of the acoustic input signal is performed in an always on mode, and wherein identifying the user that is speaking is performed after receiving a trigger phrase to activate the speech recognizer.

10. The method of claim 6 wherein classifying portions of the acoustic input signal is not performed until receiving a trigger phrase to activate the speech recognizer.

11. The method of claim 1 further comprising training the speech recognizer to recognize different users'"'"' speech signatures.

12. The method of claim 1 further comprising storing speaker-specific information for the user based on the response for use in determining additional responses.

13. The method of claim 1 wherein determining the set of potential responses comprises:

determining user preferences in the speaker-specific information; and

performing a search using the user preferences and the recognized acoustic input signal.

14. The method of claim 13

wherein the set of potential responses are ranked based on the user preferences.

15. The method of claim 9 further comprising verifying who is speaking after receiving the trigger phrase to determine if the identified user that is speaking is still speaking.

16. The method of claim 15, wherein the verifying is performed periodically.

17. The method of claim 15 wherein a second verification of who is speaking is performed when a higher security is deemed necessary.

18. The method of claim 17 wherein a manual login is not required if the second verification is performed.

19. The method of claim 1 further comprising, if it is determined that no response in the set of potential responses should be output:

refraining from outputting anything to the user.

20. A non-transitory computer readable medium having stored thereon program code executable by a processor, the program code comprising:

code that causes the processor to receive an acoustic input signal at a speech recognizer;

code that causes the processor to identify a user that is speaking based on the acoustic input signal;

code that causes the processor to recognize, via the speech recognizer, speech uttered by the user in the acoustic input signal;

code that causes the processor to determine speaker-specific information previously stored for the user;

code that causes the processor to determine a set of potential responses based on the recognized speech and the speaker-specific information for the user;

code that causes the processor to rank the set of potential responses based on one or more criteria and the speaker-specific information;

code that causes the processor to determine, for each response in the set of potential responses, whether the response should be output or should not be output based on the response'"'"'s ranking; and

if the response should be output;

code that causes the processor to select, from among a plurality of preconfigured output methods, an output method for outputting the response to the user, the selecting being based on the response'"'"'s ranking; and

code that causes the processor to output the response to the user using the selected output method.

21. A system comprising:

a processor; and

a non-transitory computer readable medium having stored thereon program code that, when executed by the processor, causes the processor to;

receive an acoustic input signal at a speech recognizer;

identify a user that is speaking based on the acoustic input signal;

recognize, via the speech recognizer, speech uttered by the user in the acoustic input signal;

determine speaker-specific information previously stored for the user;

determine a set of potential responses based on the recognized speech and the speaker-specific information for the user;

rank the set of potential responses based on one or more criteria and the speaker-specific information;

determine, for each response in the set of potential responses, whether the response should be output or should not be output based on the response'"'"'s ranking; and

if the response should be output;

select, from among a plurality of preconfigured output methods, an output method for outputting the response to the user, the selecting being based on the response'"'"'s ranking; and

output the response to the user using the selected output method.