System and method for rapid customization of speech recognition models

• US 10,726,833 B2
• Filed: 05/21/2018
• Issued: 07/28/2020
• Est. Priority Date: 03/28/2011
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Bangalore, Srinivas, Bell, Robert, Caseiro, Diamantino Antonio, Gilbert, Mazin, Haffner, Patrick
• Primary Examiner(s): Baker, Matthew H
• Application Number:
  US15/985,107
• Publication Number:
  US 20180268810A1
• Time in Patent Office: 799 Days
• CPC Class Codes: G10L 15/06 Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice G10L15/14 takes precedence
• CPC Class Codes: G10L 15/065 Adaptation
• CPC Class Codes: G10L 15/183 using context dependencies, e.g. language models
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
• CPC Class Codes: G10L 2015/0635 updating or merging of old and new templates; Mean values; Weighting
• CPC Class Codes: G10L 2015/0636 Threshold criteria for the updating
• CPC Class Codes: G10L 2015/228 of application context

Abstract

Disclosed herein are systems, methods, and non-transitory computer-readable storage media for generating domain-specific speech recognition models for a domain of interest by combining and tuning existing speech recognition models when a speech recognizer does not have access to a speech recognition model for that domain of interest and when available domain-specific data is below a minimum desired threshold to create a new domain-specific speech recognition model. A system configured to practice the method identifies a speech recognition domain and combines a set of speech recognition models, each speech recognition model of the set of speech recognition models being from a respective speech recognition domain. The system receives an amount of data specific to the speech recognition domain, wherein the amount of data is less than a minimum threshold to create a new domain-specific model, and tunes the combined speech recognition model for the speech recognition domain based on the data.

Claims

1. A method comprising:

generating a multi-domain speech recognition model comprising a combination of speech recognition models selected based on a speech pattern of a user, wherein each speech recognition model from the combination of speech recognition models is associated with a respective speech recognition domain;

receiving sample data associated with a specific speech recognition domain;

when the sample data associated with the specific speech recognition domain is more than a threshold, generating a new domain-specific speech recognition model for the specific speech recognition domain; and

when the sample data is less than the threshold, modifying the multi-domain speech recognition model by weighting components of the multi-domain speech recognition model associated with the specific speech recognition domain to recognize speech associated with the user or additional speech from the user.

2. The method of claim 1, further comprising:

identifying, based on speech from a user, the speech pattern of the user.

3. The method of claim 1, further comprising:

identifying, based on the speech pattern of the user, a plurality of speech recognition models, each speech recognition model of the plurality of speech recognition models being from a respective speech recognition domain; and

combining, via a processor, the plurality of speech recognition models, to yield the multi-domain combined speech recognition model.

4. The method of claim 1, wherein each respective speech recognition domain comprises one of business, finance, travel, medical, sports, news, politics, entertainment, and education.

5. The method of claim 3, wherein each speech recognition model in the plurality of speech recognition models performs automatic speech recognition using a list of words specific to a respective speech recognition domain of the each speech recognition model.

6. The method of claim 1, wherein the modifying of the multi-domain speech recognition model is performed on-demand in response to a request.

7. The method of claim 3, wherein the plurality of speech recognition models comprises two speech recognition models from different domains.

8. The method of claim 3, wherein the multi-domain speech recognition model and one of the plurality of speech recognition models are from different domains.

9. The method of claim 1, further comprising receiving, in addition to the sample data specific to the specific speech recognition domain, one of text, transition data, metadata, and audio, specific to the specific speech recognition domain.

10. The method of claim 1, wherein the specific speech recognition domain is specific to speech patterns of a particular user over time, wherein the speech patterns of the particular user are updated based on the speech.

11. The method of claim 1, wherein the modifying of the multi-domain speech recognition model further comprises sampling the speech.

12. The method of claim 1, further comprising recognizing additional speech using one of the new domain-specific speech recognition model and a modified multi-domain speech recognition model.

13. A system comprising:

a processor; and

a computer-readable storage medium having instruction stored which, when executed by the processor, result in the processor performing operations comprising;

generating a multi-domain speech recognition model comprising a combination of speech recognition models selected based on a speech pattern of a user, wherein each speech recognition model from the combination of speech recognition models is associated with a respective speech recognition domain;

receiving sample data associated with a specific speech recognition domain;

when the sample data associated with the specific speech recognition domain is more than a threshold, generating a new domain-specific speech recognition model for the specific speech recognition domain; and

when the sample data is less than the threshold, modifying the multi-domain speech recognition model by weighting components of the multi-domain speech recognition model associated with the specific speech recognition domain to recognize speech associated with the user or additional speech from the user.

14. The system of claim 13, wherein the computer-readable storage medium stores additional instruction stored which, when executed by the processor, result in the processor performing operations further comprising:

identifying, based on speech from a user, the speech pattern of the user.

15. The system of claim 13, wherein the computer-readable storage medium stores additional instruction stored which, when executed by the processor, result in the processor performing operations further comprising:

identifying, based on the speech pattern of the user, a plurality of speech recognition models, each speech recognition model of the plurality of speech recognition models being from a respective speech recognition domain; and

combining, via a processor, the plurality of speech recognition models, to yield the multi-domain speech recognition model.

16. A computer-readable storage device storing instructions which, when executed by a computing device, cause the computing device to perform operations comprising:

generating a multi-domain speech recognition model comprising a combination of speech recognition models selected based on a speech pattern of a user, wherein each speech recognition model from the combination of speech recognition models is associated with a respective speech recognition domain;

receiving sample data associated with a specific speech recognition domain;

when the sample data associated with the specific speech recognition domain is more than a threshold, generating a new domain-specific speech recognition model for the specific speech recognition domain;

when the sample data is less than the threshold, modifying the multi-domain speech recognition model by weighting components of the multi-domain speech recognition model associated with the specific speech recognition domain to recognize speech associated with the user or additional speech from the user.