METHODS AND APPARATUS FOR GENERATING, UPDATING AND DISTRIBUTING SPEECH RECOGNITION MODELS

• US 20120101812A1
• Filed: 12/30/2011
• Published: 04/26/2012
• 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
• Granted Patent: US 8,447,599 B2
• Time in Patent Office: Days
• US Class Current: 704/201
• 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 speech processing method, comprising:

receiving, by a speech processing system, speech data from a first computing device, the speech processing system utilizing one or more existing speech recognition models usable to recognize speech;

processing, by the speech processing system, the speech data;

based on the processing, generating, by the speech processing system, a new speech recognition model transmitting, by the speech processing system, the new speech recognition model to the first computing device; and

transmitting, by the speech processing system, the new speech recognition model to at least one second computing device different from the first computing device.

2. The method of claim 1, wherein:

generating a new speech recognition model includes generating an update to at least one of the one or more existing speech recognition models and updating the at least one of the one or more existing speech recognition models;

transmitting the new speech recognition model to the first computing device includes one of transmitting the update to the first computing device and transmitting the updated one of the one or more existing speech recognition models to the first computing device; and

transmitting the new speech recognition model to at least one second computing device includes transmitting the update to the at least one second computing device.

3. The method of claim 2, wherein the update to the at least one of the one or more existing speech recognition models includes at least one type of speech characteristic information that was not included in the one or more existing speech recognition models.

4. The method of claim 1, wherein the new speech recognition model includes at least one type of speech characteristic information that was not included in the one or more existing speech recognition models.

5. The method of claim 4, wherein the at least one type of speech characteristic information includes at least one of change in energy information and change in amplitude information.

6. The method of claim 1, wherein the speech processing system is physically remote from the first computing device and the at least one second computing device.

7. The method of claim 1, wherein the first computing device is not configured to perform speech recognition modeling.

8. The method of claim 1, wherein the at least one second computing device is not configured to perform speech recognition modeling.

9. The method of claim 1, wherein the update to the first computing device has a different recognition model format than a previous speech recognition model provided on the first computing device.

10. The method of claim 1, wherein the one or more existing speech recognition models include model-type specific and model-type independent speech recognition models.

11. The method of claim 1, wherein the one or more existing speech recognition models include speaker dependent and speaker independent speech recognition models.

12. The method of claim 1, wherein the one or more existing speech recognition models include application specific and application independent speech recognition models.

13. 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 speech data from a first computing device, the speech processing system utilizing one or more existing speech recognition models usable to recognize speech;

processing the speech data, based on the processing, generating, by the speech processing system, a new speech recognition model;

transmitting the new speech recognition model to the first remote device; and

transmitting the new speech recognition model to at least one second computing device different from the first computing device.

14. The non-transitory computer-readable medium of claim 13, wherein:

generating a new speech recognition model includes generating an update to at least one of the one or more existing speech recognition models and updating the at least one of the one or more existing speech recognition models;

transmitting the new speech recognition model to the first remote device includes one of transmitting the update to the first remote device and transmitting the updated one of the one or more existing speech recognition models to the first remote device; and

transmitting the new speech recognition model to at least one second computing device includes transmitting the update to the at least one second computing device.

15. The non-transitory computer-readable medium of claim 14, wherein the update to the at least one of the one or more existing speech recognition models includes at least one type of speech characteristic information that was not included in the one or more existing speech recognition models.

16. The non-transitory computer-readable medium of claim 13, wherein the new speech recognition model includes at least one type of speech characteristic information that was not included in the one or more existing speech recognition models.

17. A speech processing 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 computers to perform operations comprising;

receiving speech data from a first computing device, the speech processing system utilizing one or more existing speech recognition models usable to recognize speech;

processing the speech data;

based on the processing, generating a new speech recognition model transmitting, by the speech processing system, the new speech recognition model to the first computing device; and

transmitting the new speech recognition model to at least one second computing device different from the first computing device.

18. The system of claim 17, wherein:

generating a new speech recognition model includes generating an update to at least one of the one or more existing speech recognition models and updating the at least one of the one or more existing speech recognition models;

transmitting the new speech recognition model to the first computing device includes one of transmitting the update to the first computing device and transmitting the updated one of the one or more existing speech recognition models to the first computing device; and

transmitting the new speech recognition model to at least one second computing device includes transmitting the update to the at least one second computing device.

19. The system of claim 18, wherein the update to the at least one of the one or more existing speech recognition models includes at least one type of speech characteristic information that was not included in the one or more existing speech recognition models.

20. The system of claim 17, wherein the new speech recognition model includes at least one type of speech characteristic information that was not included in the one or more existing speech recognition models.