Voice recognition terminal, server, method of controlling server, voice recognition system, non-transitory storage medium storing program for controlling voice recognition terminal, and non-transitory storage medium storing program for controlling server

• US 9,626,964 B2
• Filed: 11/24/2014
• Issued: 04/18/2017
• Est. Priority Date: 11/27/2013
• Status: Active Grant

• Current Assignee: Sharp Kabushiki Kaisha (Hon Hai Precision Industry Co., Ltd.)
• Original Assignee: Sharp Kabushiki Kaisha (Hon Hai Precision Industry Co., Ltd.)
• Inventors: Iwano, Yuri, Hirata, Masafumi, Tojima, Akira
• Primary Examiner(s): SINGH, SATWANT K
• Application Number:
  US14/552,109
• Publication Number:
  US 20150149175A1
• Time in Patent Office: 876 Days
• Field of Search: None
• 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

Abstract

A voice recognition terminal is provided to be able to communicate with a server capable of voice recognition for recognizing voice, and includes a voice input acceptance portion accepting voice input from a user, a voice recognition portion carrying out voice recognition of the voice input accepted, a response processing execution portion performing processing for responding to the user based on a result of voice recognition of the voice input accepted, and a communication portion transmitting the voice input accepted by the voice input acceptance portion to the server and receiving a result of voice recognition in the server. The response processing execution portion performs the processing for responding to the user based on the result of voice recognition determined as more suitable, of the result of voice recognition by the voice recognition portion and the result of voice recognition received from the server.

Claims

1. A voice recognition terminal provided to be able to communicate with a server capable of voice recognition for recognizing voice and calculating reliability indicating correctness of recognition, comprising:

a voice input acceptance portion accepting voice input from a user;

a voice recognition portion carrying out voice recognition of the voice input accepted by said voice input acceptance portion;

a response processing execution portion performing processing for responding to said user based on a result of voice recognition of the voice input accepted by said voice input acceptance portion; and

a communication portion transmitting the voice input accepted by said voice input acceptance portion to said server and receiving a result of voice recognition in said server,

said response processing execution portion performing the processing for responding to said user based on the result of voice recognition determined as more suitable, of the result of voice recognition by said voice recognition portion and the result of voice recognition received from said server, wherein

said voice recognition portion recognizes said voice and calculates reliability indicating correctness of recognition, and

said response processing execution portion further performs the processing for responding to said user based on the result of voice recognition obtained later between the result of voice recognition received from said server and the result of voice recognition carried out by the voice recognition portion, when reliability included in the later obtained result of voice recognition is higher than reliability included in the result of voice recognition obtained earlier, in connection with reliability included in the result of voice recognition by said voice recognition portion, such that smooth communication can be carried out with the user without imposing stress to the user.

2. The voice recognition terminal according to claim 1, wherein

said response processing execution portion further performs the processing for responding to said user based on the later obtained result of voice recognition when reliability included in the earlier obtained result of voice recognition is equal to or lower than prescribed reliability and when the reliability included in the later obtained result of voice recognition is higher than the reliability included in the earlier obtained result of voice recognition, in connection with reliability included in the result of voice recognition by said voice recognition portion.

3. The voice recognition terminal according to claim 1, further comprising an operation mode switching portion switching between a server mode in which said voice recognition terminal operates in accordance with an instruction from said server and a normal mode in which said voice recognition terminal operates by selectively making use of the instruction from said server, based on the result of voice recognition of the voice input provided to said voice input acceptance portion.

4. The voice recognition terminal according to claim 3, wherein

said response processing execution portion invalidates the result of voice recognition by said voice recognition portion in said server mode.

5. A server provided to be able to communicate with a voice recognition terminal capable of voice recognition for recognizing voice and calculating reliability indicating correctness of recognition, comprising:

a voice input reception portion receiving voice input from a user through said voice recognition terminal;

a voice recognition portion carrying out voice recognition of the voice input received by said voice input reception portion; and

a response processing execution instruction portion instructing said voice recognition terminal to perform processing for responding to said user based on a result of voice recognition of the voice input received by said voice input reception portion,

said response processing execution instruction portion instructing said voice recognition terminal to switch from a normal mode in which said voice recognition terminal operates by selectively making use of an instruction from said server to a server mode instructing said voice recognition terminal to operate in accordance with the instruction from said server, wherein

said voice recognition portion recognizes said voice and calculates reliability indicating correctness of recognition, and

said response processing execution portion further performs the processing for responding to said user based on the result of voice recognition obtained later between a result of voice recognition carried out by the voice recognition terminal and the result of voice recognition carried out by the voice recognition portion, when reliability included in the later obtained result of voice recognition is higher than reliability included in the result of voice recognition obtained earlier, in connection with reliability included in the result of voice recognition by said voice recognition portion, such that smooth communication can be carried out with the user without imposing stress to the user.

6. The server according to claim 5, wherein

said response processing execution instruction portion

determines whether the result of voice recognition of the voice input accepted by said voice input reception portion is a request for a query about prescribed information from said user,

determines whether data for obtaining the prescribed information has been registered when it is determined that the result is the request for the query about said prescribed information, and

indicates execution of response processing inviting said user to input said data when the data for obtaining the prescribed information has not been registered, based on a result of determination.

7. A method of controlling a server provided to be able to communicate with a voice recognition terminal capable of voice recognition for recognizing voice and calculating reliability indicating correctness of recognition, comprising:

receiving voice input from a user through said voice recognition terminal;

carrying out voice recognition of the received voice input; and

instructing said voice recognition terminal to perform processing for responding to said user based on a result of voice recognition of the received voice input,

instructing said voice recognition terminal to perform processing for responding including indicating switching from a normal mode in which said voice recognition terminal operates by selectively making use of an instruction from said server to a server mode instructing said voice recognition terminal to operate in accordance with the instruction from said server, wherein

said step of carrying out voice recognition includes recognizing, said voice and calculating reliability indicating correctness of recognition, and

said step of instructing said voice recognition terminal includes further performing the processing for responding to said user based on the result of voice recognition obtained later between a result of voice recognition carried out by the voice recognition terminal and the result of voice recognition carried out in the step of carrying out voice recognition, when reliability included in the later obtained result of voice recognition is higher than reliability included in the result of voice recognition obtained earlier, in connection with reliability included in the result of voice recognition by said step of carrying out voice recognition, such that smooth communication can be carried out with the user without imposing stress to the user.

8. A voice recognition system, comprising:

a server capable of voice recognition for recognizing voice and calculating reliability indicating correctness of recognition; and

a voice recognition terminal provided to be able to communicate with said server,

said voice recognition terminal including

a voice input acceptance portion accepting voice input from a user,

a voice recognition portion carrying out voice recognition of the voice input accepted by said voice input acceptance portion,

a response processing execution portion performing processing for responding to said user based on a result of voice recognition of the voice input accepted by said voice input acceptance portion, and

a communication portion transmitting the voice input accepted by said voice input acceptance portion to said server and receiving a result of voice recognition in said server, and

said response processing execution portion performing the processing for responding to said user based on the result of voice recognition determined as more suitable, of the result of voice recognition by said voice recognition portion and the result of voice recognition received from said server, wherein

said voice recognition portion recognizes said voice and calculates reliability indicating correctness of recognition, and

said response processing execution portion further performs the processing for responding to said user based on the result of voice recognition obtained later between a result of voice recognition carried out by the server and the result of voice recognition carried out by the voice recognition portion, when reliability included in the later obtained result of voice recognition is higher than reliability included in the result of voice recognition obtained earlier, in connection with reliability included in the result of voice recognition by said voice recognition portion, such that smooth communication can be carried out with the user without imposing stress to the user.

9. A method of controlling a voice recognition terminal provided to be able to communicate with a server capable of voice recognition for recognizing voice and calculating reliability indicating correctness of recognition, comprising:

accepting voice input from a user;

carrying out voice recognition of the accepted voice input;

performing processing for responding to said user based on a result of voice recognition of the accepted voice input; and

transmitting the accepted voice input to said server and receiving a result of voice recognition in said server,

performing processing for responding including performing processing for responding to said user based on the result of voice recognition determined as more suitable, of the result of voice recognition of said accepted voice input and the result of voice recognition received from said server, wherein

said step of carrying out voice recognition includes recognizing said voice and calculating reliability indicating correctness of recognition, and

said step of performing processing for responding includes further performing processing for responding to said user based on the result of voice recognition obtained later between the result of voice recognition carried out b the server and the result of voice recognition carried out by the voice recognition portion, when reliability included in the later obtained result of voice recognition is higher than reliability included in the result of voice recognition obtained earlier, in connection with reliability included in the result of voice recognition by said step of carrying out voice recognition, such that smooth communication can be carried out with the user without imposing stress to the user.

10. The method of controlling a voice recognition terminal according to claim 9, wherein

performing said processing for responding includes further performing the processing for responding to said user based on the later obtained result of voice recognition when reliability included in the earlier obtained result of voice recognition is equal to or lower than prescribed reliability and when reliability included in the later obtained result of voice recognition is higher than reliability included in the earlier obtained result of voice recognition, in connection with reliability included in the result of voice recognition of said accepted voice input.

11. The method of controlling a voice recognition terminal according to claim 9, further comprising switching between a server mode in which said voice recognition terminal operates in accordance with an instruction from said server and a normal mode in which said voice recognition terminal operates by selectively making use of the instruction from said server, based on the result of voice recognition of said accepted voice input.

12. The method of controlling a voice recognition terminal according to claim 11, wherein

performing said processing for responding includes invalidating the result of voice recognition of said accepted voice input in said server mode.