Multilevel speech recognition for candidate application group using first and second speech commands

Patent metadata

Field	Value
Publication number	US-10043520-B2
Application number	US-201414558479-A
Country	US
Kind code	B2
Filing date	Dec 2, 2014
Priority date	Jul 9, 2014
Publication date	Aug 7, 2018
Grant date	Aug 7, 2018

Abstract

A multilevel speech recognition method and an apparatus performing the method are disclosed. The method includes receiving a first speech command from a user through a speech interface, and extracting a keyword from the first speech command. The method also includes providing a candidate application group of a category providing a service associated with the keyword, and processing a second speech command from the user associated with an application selected from the candidate application group.

First claim

What is claimed is:

1. A multilevel speech recognition method, the method comprising:
extracting a keyword from a first speech command received from a user;
determining whether the extracted keyword from the first speech command of the user includes an application indication for an application using an application speech recognition model;
in response to the extracted keyword from the first speech command being determined to not include the application indication, providing a candidate application group of a category providing a service determined to be associated with the extracted keyword, and identifying a selected application selected by the user from the provided candidate application group using a category speech recognition model based on a set of words associated with the category; and
processing a second speech command, from the user, for the selected application selected from the provided candidate application group, using the application speech recognition model, based on a set of words associated with the selected application,
wherein the application speech recognition model is updated based on at least one of a voice, a pronunciation, or an accent of the user obtained from the second speech command, and a word frequently received from the user included in the second speech command, the word being associated with the selected application.

2. The method of claim 1 , wherein the category is determined based on a use pattern of the user obtained from the first speech command and the second speech command.

3. The method of claim 1 , wherein the category speech recognition model is updated based on a characteristic and a use pattern of the user obtained from the first speech command.

4. A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the method of claim 1 .

5. The method of claim 1 , wherein the application speech recognition model is updated based on at least one of a voice, a pronunciation, or an accent of the user obtained from speech commands, and a word frequently received from the user included in the speech commands, and the processing of the second speech command uses the updated application speech recognition model.

6. A multilevel speech recognition method, the method comprising:
determining, using a processor, whether a speech command of a user includes an application indication for an application by processing the speech command using an application speech recognition model,
wherein the application speech recognition model is updated based on at least one of a voice, a pronunciation, or an accent of the user obtained from the speech command, and a word frequently received from the user included in speech commands, when the word is associated with the application indication;
executing, at the processor, the application in response to the speech command being determined to include the application indication; and
in response to the speech command being determined to not include the application indication, identifying a selected application selected by the user from a provided candidate application group of a category determined to be associated with the word using a category speech recognition model based on a set of words associated with the category, and executing the selected application selected from the provided candidate application group, the selected application providing a service associated with the speech command.

7. The method of claim 6 , wherein the processing comprises:
processing the speech command using the application speech recognition model, comprising an acoustic model and a linguistic model, based on a set of words associated with the selected application.

8. The method of claim 6 , further comprising:
processing, using the category speech recognition model, comprising an acoustic model and a linguistic model based on words associated with a category, and the application speech recognition model, comprising an acoustic model and a linguistic model, based on words associated with the selected application, to provide a result identifying the speech command to a user terminal.

9. The method of claim 8 , further comprising:
updating the category speech recognition model and the application speech recognition model as the processor receives additional speech commands.

10. The method of claim 6 , wherein the determining comprises
recognizing the word frequently received from the user through a microphone, and
recognizing the word frequently received from the user as being associated with the application.

11. The method of claim 6 , wherein the determining comprises
updating the selected application based on the updated application speech recognition model.

12. A non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the method of claim 6 .

13. A user terminal, comprising:
a processor, configured to
extract a keyword from a first speech command received from a user,
determine whether the extracted keyword from the first speech command of the user includes an application indication for an application using an application speech recognition model;
in response to the extracted keyword from the first speech command being determined to not include the application indication, provide a candidate application group of a category providing a service determined to be associated with the extracted keyword, and identify a selected application selected by the user from the provided candidate application group using a category speech recognition model based on a set of words associated with the category; and
process a second speech command, from the user, for the selected application selected from the provided candidate application group, using the application speech recognition model, based on a set of words associated with the selected application, wherein
the application speech recognition model is updated based on at least one of a voice, a pronunciation, or an accent of the user obtained from the second speech command, and a word frequently received from the user included in the second speech command, the word being associated with the selected application.

14. The user terminal of claim 13 , wherein the processor is further configured to:
recognize the selected application selected using an acoustic model and a linguistic model based on the set of words associated with the category, and
process the second speech command using an acoustic model and a linguistic model, based on the set of words associated with the selected application.

15. The user terminal of claim 13 , wherein the processor is further configured to update the category speech recognition model and the application speech recognition model as the processor receives additional speech commands.

16. The user terminal of claim 13 , wherein the processor is further configured to analyze a use pattern and a voice of the user based on a word included in the first speech command and update the category speech recognition model and the application speech recognition model based on a result of the analyzing.

17. The user terminal of claim 13 , further comprising:
a speech interface, operatively connected to the processor, wherein the processor is configured to receive the first speech command of the user through the speech interface.

Assignees

• Samsung Electronics Co Ltd

Inventors

• Mun Minyoung
• Choi Youngsang

Classifications

• G10L2015/223

  Execution procedure of a spoken command · CPC title

• G10L15/1822

  Parsing for meaning understanding · CPC title

• G10L15/32 Primary

  Multiple recognisers used in sequence or in parallel; Score combination systems therefor, e.g. voting systems · CPC title