Internal and external speech recognition use with a mobile communication facility

• US 8,949,130 B2
• Filed: 10/21/2009
• Issued: 02/03/2015
• Est. Priority Date: 03/07/2007
• Status: Expired due to Fees

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: Vlingo Corporation (Microsoft Corporation)
• Inventors: Phillips, Michael S.
• Primary Examiner(s): COLUCCI, MICHAEL C
• Application Number:
  US12/603,446
• Publication Number:
  US 20100106497A1
• Time in Patent Office: 1,931 Days
• Field of Search: 704/270.1, 704/275, 704/270, 704/257, 704/256.2, 704/255, 704/254, 704/244, 704/235, 704/231, 715/830, 715/835, 706/20, 455/557, 717/168, 382/190
• US Class Current: 704/270
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
• CPC Class Codes: G10L 2015/228 of application context

Abstract

In embodiments of the present invention improved capabilities are described for a user interacting with a mobile communication facility, where speech presented by the user is recorded using a mobile communication facility resident capture facility. The recorded speech may be recognized using an external speech recognition facility to produce an external output and a resident speech recognition facility to produce an internal output, where at least one of the external output and the internal output may be selected based on a criteria.

Claims

1. A method of interacting with a mobile communication facility comprising:

recording, at a mobile communication facility resident capture facility, speech of a user;

attempting to recognize the speech, at an external speech recognition facility, to produce a first recognized speech result, wherein one or more language models are adapted based on information about the user including a usage history of the user;

recognizing the speech, at a resident speech recognition facility, to produce a second recognized speech result, wherein adaptation is based upon, at least in part, a device specific identifier transmitted from the resident speech recognition facility to the external speech recognition facility;

selecting, based on a criteria, a selected output that includes at least one of the first recognized speech result and the second recognized speech result, wherein attempting to recognize the speech using the external speech recognition facility is aborted prior to conclusion when the second recognized speech result is selected; and

using the selected output to perform a function on the mobile communication facility.

2. The method of claim 1 wherein the criteria is a confidence score.

3. The method of claim 1 wherein the criteria is based on an amount of time to receive the first recognized speech result.

4. The method of claim 1, wherein the second recognized speech result is selected based on the criteria.

5. The method of claim 4, wherein the resident speech recognition facility is constrained by a grammar and the external speech recognition facility is not constrained by the grammar.

6. The method of claim 4, wherein the criteria is based in part on information from the resident speech recognition facility.

7. The method of claim 6, wherein the information from the resident speech recognition facility includes a score which indicates a confidence that results provided by the resident speech recognition facility correctly represents the speech.

8. The method of claim 4, wherein the resident speech recognition facility is adapted based on information about the user.

9. The method of claim 8, where the information about the user includes a usage history of the user.

10. The method of claim 8, wherein this adaptation is performed externally.

11. The method of claim 8, wherein this adaptation is performed on the mobile communication facility.

12. The method of claim 4, wherein selecting the selected output is based in part on an amount of time taken by the external speech recognition facility and on information from the resident speech recognition facility, when the selected output includes the second recognized speech result.

13. The method of claim 12, wherein the resident speech recognition facility is constrained by a grammar and the external speech recognition facility is not constrained by the grammar.

14. The method of claim 12, wherein the resident speech recognition facility is adapted based on information about the user.

15. The method of claim 14, where the information about the user includes a usage history of the user.

16. The method of claim 14, wherein this adaptation is performed externally.

17. The method of claim 14, wherein this adaptation is performed on the mobile communication facility.

18. The method of claim 1, wherein this adaptation is performed externally.

19. The method of claim 1, wherein this adaptation is performed on the mobile communication facility.

20. A method of interacting with a mobile communication facility comprising:

recording, at a mobile communication facility resident capture facility, speech of a user;

communicating the speech to an external speech recognition facility to attempt to produce a first recognized speech result, wherein one or more language models are adapted based on information about the user including a usage history of the user;

recognizing the speech, at a resident speech recognition facility, to produce a second recognized speech result, wherein adaptation is based upon, at least in part, a device specific identifier transmitted from the resident speech recognition facility to the external speech recognition facility;

selecting the second recognized speech result based on a confidence score criteria, wherein the confidence score criteria is based in part on information from the resident speech recognition facility; and

using the second recognized speech result to perform a function on the mobile communication facility.

21. A system for interacting with a mobile communication facility comprising:

a mobile communication facility resident capture facility for recording speech of a user;

an external speech recognition facility for attempting to recognize the speech to produce a first recognized speech result, wherein one or more language models are adapted based on information about the user including a usage history of the user;

a resident speech recognition facility for recognizing the speech to produce a second recognized speech result, wherein adaptation is based upon, at least in part, a device specific identifier transmitted from the resident speech recognition facility to the external speech recognition facility; and

a resident selecting facility to select a selected output that includes at least one of the first recognized speech result and the second recognized speech result based on a criteria, wherein the selected output by the resident selecting facility performs a function on the mobile communication facility, wherein the second recognized speech result is selected based on the criteria and wherein the second recognized speech result is selected based in part on an amount of time taken by the external speech recognition facility.

22. The system of claim 21, wherein a first amount of time taken by the external speech recognition facility is greater than a second amount of time for the resident speech recognition facility because the resident speech recognition facility is constrained by a grammar and the external speech recognition facility is not constrained by the grammar.

23. The system of claim 21, wherein selecting the selected output is based in part on an amount of time taken by the external speech recognition facility and on information from the resident speech recognition facility, when the selected output includes the second recognized speech result.