Methods and apparatus for generating, updating and distributing speech recognition models

• US 8,494,848 B2
• Filed: 09/13/2012
• Issued: 07/23/2013
• Est. Priority Date: 11/30/2000
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Reding, Craig, Levas, Suzi
• Primary Examiner(s): Han, Qi
• Application Number:
  US13/614,982
• Publication Number:
  US 20130013298A1
• Time in Patent Office: 313 Days
• Field of Search: 704/201, 704/231, 704/235, 704/251, 704/270
• US Class Current: 704/231
• CPC Class Codes: G10L 15/00 Speech recognition G10L17/00 takes precedence
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications

Abstract

Techniques for generating, distributing, and using speech recognition models are described. A shared speech processing facility is used to support speech recognition for a wide variety of devices with limited capabilities including business computer systems, personal data assistants, etc., which are coupled to the speech processing facility via a communications channel, e.g., the Internet. Devices with audio capture capability record and transmit to the speech processing facility, via the Internet, digitized speech and receive speech processing services, e.g., speech recognition model generation and/or speech recognition services, in response. The Internet is used to return speech recognition models and/or information identifying recognized words or phrases. Thus, the speech processing facility can be used to provide speech recognition capabilities to devices without such capabilities and/or to augment a device'"'"'s speech processing capability. Voice dialing, telephone control and/or other services are provided by the speech processing facility in response to speech recognition results.

Claims

1. A computer-implemented method comprising:

receiving a request to update an existing speech recognition model for recognizing a particular term by a speech recognizer on a mobile device, wherein the request references a textual representation of the particular term and includes a recording of a user of the mobile device speaking the particular term or a set of features extracted from the recording, wherein the request specifies an updated speech recognition model type to be generated, and wherein the updated speech recognition type is a different speech recognition model type than the speech recognition model type of the existing speech recognition model;

generating the updated speech recognition model for recognizing the particular term based on the request; and

transmitting the updated speech recognition model for recognizing the particular term to the mobile device, as a replacement for the existing speech recognition model for recognizing the particular term.

2. The method of claim 1, wherein the request further includes an identifier of a user associated with the mobile device.

3. The method of claim 1, wherein the request further includes speech data associated with the term.

4. The method of claim 1, wherein the request further includes feature data extracted, by the mobile device, from speech data associated with the term.

5. The method of claim 1, further comprising:

receiving a second recording of the user of the mobile device speaking the particular term or a set of features extracted from the second recording; and

recognizing the particular term in the second recording based on the updated speech recognition model.

6. The method of claim 1, comprising:

augmenting a training corpus with the term, wherein the updated speech recognition model is generated using the augmented training corpus.

7. The method of claim 1, wherein the updated speech recognition model is a speaker-dependent recognition model.

8. The method of claim 1, comprising generating one or more speaker-independent speech recognition models using the term.

9. A system comprising:

one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising;

receiving a request to update an existing speech recognition model for recognizing a particular term by a speech recognizer on a mobile device, wherein the request references a textual representation of the particular term and includes a recording of a user of the mobile device speaking the particular term or a set of features extracted from the recording, wherein the request specifies an updated speech recognition model type to be generated, and wherein the updated speech recognition type is a different speech recognition model type than the speech recognition model type of the existing speech recognition model;

generating the updated speech recognition model for recognizing the particular term based on the request; and

transmitting the updated speech recognition model for recognizing the particular term to the mobile device, as a replacement for the existing speech recognition model for recognizing the particular term.

10. The system of claim 9, wherein the request further includes an identifier of a user associated with the mobile device.

11. The system of claim 9, wherein the request further includes speech data associated with the term.

12. The system of claim 9, wherein the request further includes feature data extracted, by the mobile device, from speech data associated with the term.

13. The system of claim 9, the operations further comprising:

receiving a second recording of the user of the mobile device speaking the particular term or a set of features extracted from the second recording; and

recognizing the particular term in the second recording based on the updated speech recognition model.

14. The system of claim 9, wherein the operations comprise:

augmenting a training corpus with the term, wherein the updated existing speech recognition model is generated using the augmented training corpus.

15. The system of claim 9, wherein the updated speech recognition model is a speaker-dependent recognition model.

16. The system of claim 9, wherein the operations comprise generating one or more speaker-independent speech recognition models using the term.

17. A non-transitory computer-readable medium storing software comprising instructions executable by one or more computers which, upon such execution, cause the one or more computers to perform operations comprising:

receiving a request to update an existing speech recognition model for recognizing a particular term by a speech recognizer on a mobile device, wherein the request references a textual representation of the particular term and includes a recording of a user of the mobile device speaking the particular term or a set of features extracted from the recording, wherein the request specifies an updated speech recognition model type to be generated, and wherein the updated speech recognition type is a different speech recognition model type than the speech recognition model type of the existing speech recognition model;

generating the updated speech recognition model for recognizing the particular term based on the request; and

transmitting the updated speech recognition model for recognizing the particular term to the mobile device, as a replacement for the existing speech recognition model for recognizing the particular term.

18. The medium of claim 17, wherein the operations comprise generating one or more speaker-independent speech recognition models using the term.

19. The medium of claim 17, the operations further comprising:

receiving a second recording of the user of the mobile device speaking the particular term or a set of features extracted from the second recording; and

recognizing the particular term in the second recording based on the updated speech recognition model.