VOICE RECOGNITION DEVICE AND METHOD OF CONTROLLING SAME

• US 20150310855A1
• Filed: 12/09/2013
• Published: 10/29/2015
• Est. Priority Date: 12/07/2012
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: BAK, Eun-Sang, KIM, Kyung-duk, KIM, Myung-jae, LIU, Yu, RYU, Seong-han, LEE, Geun-bae
• Granted Patent: US 9,953,645 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G10L 15/1822 Parsing for meaning understanding
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/228 of application context

Abstract

A voice recognition device includes an extractor configured to extract at least one of a first utterance element indicating an execution command and a second utterance element indicating a subject, from a user'"'"'s voice utterance, a domain determiner configured to determine a current domain to provide response information regarding the voice utterance based on the first and the second utterance elements, and a controller configured to determine a candidate conversation frame to provide the response information regarding the voice utterance on at least one of the current domain and a previous domain based on a conversation state of the current domain, wherein the previous domain is determined from a previous voice utterance of the user.

Claims

1. A voice recognition device, comprising:

an extractor configured to extract at least one of a first utterance element indicating an execution command and a second utterance element indicating a subject, from a user'"'"'s voice utterance;

a domain determiner configured to determine a current domain to provide response information regarding the voice utterance based on the first and the second utterance elements; and

a controller configured to determine a candidate conversation frame to provide the response information regarding the voice utterance on at least one of the current domain and a previous domain based on a conversation state of the current domain,

wherein the previous domain is determined from a previous voice utterance of the user.

2. The voice recognition device of claim 1, wherein the domain determiner is further configured to determine the current domain to provide response information regarding the voice utterance based on a main action and parameters corresponding to the first and the second utterance elements extracted by the extractor.

3. The voice recognition device of claim 2, wherein the controller is further configured to determine whether or not the current domain and the previous domain are the same and whether the conversation context is converted from the current conversation frame and previous conversation frame generated regarding the previous domain, and determine a candidate conversation frame to provide response information regarding the voice utterance on at least one of the current domain and the previous domain.

4. The voice recognition device of claim 3, wherein, when the current domain and the previous domain are the same and when the conversation context is not converted between the current domain and the previous domain, the controller is further configured to determine the candidate conversation frame regarding the current conversation frame based on the previous conversation frame.

5. The voice recognition device of claim 3, wherein, when the current domain and the previous domain are different and when the conversation context is not converted between the current domain and the previous domain, the controller is further configured to determine the candidate conversation frame regarding the current conversation frame on the previous and the current domains based on the previous conversation frame.

6. The voice recognition device of claim 3, wherein, when the current domain and the previous domain are the same and when the conversation context is converted between the two domains, the controller is further configured to determine the candidate conversation frame from at least one of the current conversation frame and an initialized conversation frame which is initialized in relation with the current conversation frame on the previous domain.

7. The voice recognition device of claim 3, wherein, when the current domain and the previous domain are different and when the conversation context is converted between the two domains, the controller is further configured to determine at least one of a candidate conversation frame regarding the current conversation frame based on the previous conversation frame, a candidate conversation frame regarding the current conversation frame on the previous domain, and a candidate conversation frame regarding the initialized conversation frame initialized in relation with the current conversation frame on the current domain.

8. The voice recognition device of claim 1, further comprising:

a memory configured to store conversation example information related with the previous conversation frame matched for each domain, and counting information relating to the frequency regarding the voice utterance related with the conversation example information.

9. The voice recognition device of claim 8, wherein the controller is further configured to determine ranking of the candidate conversation frame based on the counting information matched with the conversation example information for each of the at least one previous conversation frame stored in the memory, and provide response information regarding the candidate conversation frame in the order of higher-ranking candidate conversation frame.

10. The voice recognition device of claim 8, wherein the memory is further configured to store indexing information to index at least one utterance element included in the conversation example information for each of at least one previous conversation frame, and

the controller is further configured to provide response information regarding the candidate conversation frame in the order of larger number of the indexing information among the candidate conversation frames to provide response information regarding the voice utterance by considering the indexing information of the conversation example information for each of the at least one previous conversation frame stored in the storage.

11. A method of controlling a voice recognition device, comprising:

extracting at least one of a first utterance element indicating an execution command and a second utterance element indicating a subject from a voice utterance of a user;

determining a current domain to provide a response information regarding the voice utterance based on the first and the second utterance elements;

determining a candidate conversation frame to provide the response information regarding the voice utterance on at least one of a current domain and a previous domain, based on a conversation state on the current domain and the previous domain, wherein the previous domain is determined from a previous voice utterance of the user; and

providing the response information regarding the voice utterance based on the candidate conversation frame.

12. The method of claim 11, wherein the determining comprises determining the current domain to provide response information regarding the voice utterance based on a main action and parameters corresponding to the extracted first and second utterance elements.

13. The method of claim 12, wherein the providing comprises determining whether the current domain and the previous domain are the same and whether a conversation context is converted from the current conversation frame and a previous conversation frame generated regarding the previous domain, and determines a candidate conversation frame to provide response information regarding the voice utterance on at least one of the current and the previous domains.

14. The method of claim 13, wherein, when the current domain and the previous domain are the same and when the conversation context is not converted between the two domains, the providing comprises determining a candidate conversation frame regarding the current conversation frame based on the previous conversation frame.

15. The method of claim 13, wherein, when the current domain and the previous domain are different and when the conversation context is not converted between the two domains, the providing comprises determining a candidate conversation frame regarding the current conversation frame on the previous and the current domains based on the previous conversation frame.

16. The method of claim 13, wherein, when the current domain and the previous domain are the same and when the conversation context is converted between the two domains, the providing comprises determining a candidate conversation frame from at least one of the current conversation frame and an initialized conversation frame which is initialized in relation with the current conversation frame on the previous domain.

17. The method of claim 13, wherein, when the current domain and the previous domain are different and when the conversation context is converted between the two domains, the providing comprises determining at least one of a candidate conversation frame regarding the current conversation frame based on the previous conversation frame, a candidate conversation frame regarding the current conversation frame on the previous domain, and a candidate conversation frame regarding the initialized conversation frame initialized in relation with the current conversation frame on the current domain.

18. The method of claim 11, further comprising:

matching and storing conversation example information related with the previous conversation frame matched for each domain, and counting information relating to the frequency regarding the voice utterance related with the conversation example information.

19. The method of claim 18, further comprising:

determining ranking of the candidate conversation frame based on the counting information matched with the conversation example information for each of the at least one previous conversation frame;

and providing response information regarding the candidate conversation frame in the order of higher-ranking candidate conversation frame.

20. The method of claim 18, further comprising:

storing indexing information to index at least one utterance element included in the conversation example information for each of at least one previous conversation frame; and

providing response information regarding the candidate conversation frame in the order of larger number of the indexing information among the candidate conversation frames to provide response information regarding the voice utterance by considering the indexing information of the conversation example information for each of the at least one previous conversation frame.