METHOD, INTERACTION DEVICE, SERVER, AND SYSTEM FOR SPEECH RECOGNITION

• US 20150310864A1
• Filed: 07/06/2015
• Published: 10/29/2015
• Est. Priority Date: 01/06/2013
• Status: Active Grant

• Current Assignee: Huawei Technologies Co., Ltd. (Huawei Investment & Holding Co., Ltd.)
• Original Assignee: Huawei Technologies Co., Ltd. (Huawei Investment & Holding Co., Ltd.)
• Inventors: LI, Shanfu, Dong, Mingjie
• Granted Patent: US 10,229,684 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G10L 15/1822 Parsing for meaning understanding
• CPC Class Codes: G10L 15/26 Speech to text systems G10L15/08 takes precedence
• CPC Class Codes: G10L 2015/223 Execution procedure of a spoken command
• CPC Class Codes: H04M 1/72403 with means for local support of applications that increase the functionality
• CPC Class Codes: H04M 2250/74 with voice recognition means

Abstract

Embodiments of the present invention provide a method, an apparatus, and a system for speech recognition. a third-party application corresponding to a speech signal of a user can be determined according to the speech signal and by means of semantic analysis; and third-party application registry information is searched for and a third-party program is started, so that the user does not need to tap the third-party application to start the corresponding program, thereby providing more intelligent service for the user and facilitating use for the user.

Claims

1. A speech recognition method, wherein:

parsing one or more first texts to obtain first target semantics, wherein the first text is obtained through conversion from a first speech signal;

determining, according to the acquired first target semantics, a third-party application object corresponding to the first target semantics, wherein a third-party application that is associated with the third-party application object is a program that is not authorized to start by means of speech before delivery of a device; and

acquiring, from a third-party application registry, a third-party application that is associated with the determined third-party application object and starting the third-party application.

2. The method according to claim 1, wherein before the parsing the first text to obtain first target semantics, the method further comprises:

acquiring first speech signals and converting the first speech signals to obtain the one or more first texts.

3. The method according to claim 1, wherein the converting the first speech signals to obtain more first texts comprises:

converting the first speech signals to obtain at least two first texts; and

correspondingly, the parsing the first text to obtain first target semantics comprises;

scoring semantics of each first text of the at least two first texts according to a predetermined scoring rule and according to naturalness and coherence of the semantics of the at least two first texts, wherein a higher score represents better naturalness and coherence of the semantics; and

acquiring, from the semantics of the at least two first texts, semantics with a highest score and of the first text as the first target semantics.

4. The method according to claim 3, wherein the acquiring, from the semantics of the at least two first texts, semantics with a highest score and of the first text as the first target semantics comprises:

acquiring, from the semantics of the at least two first texts, semantics, with a highest score, of the first text and exceeds a preset threshold as the first target semantics.

5. The method according to claim 1, wherein the determining, according to the acquired first target semantics, a third-party application object corresponding to the first target semantics specifically comprises:

determining, according to the first target semantics, all third-party application objects corresponding to the first target semantics;

if the number of the third-party application objects is greater than or equal to two, acquiring information about all the third-party application objects that are associated with the first target semantics;

sending prompt information, wherein the prompt information comprises the information about the third-party application objects and is used to prompt a user to further send, according to the information about the third-party application objects, a selection instruction, so as to select a target third-party application object from all the third-party application objects that are associated with the first target semantics; and

receiving the selection instruction, and selecting a corresponding target third-party application object according to the selection instruction; and

correspondingly, the acquiring, from a third-party application registry, a third-party application that is associated with the determined third-party application object and starting the third-party application specifically comprises;

acquiring, from the third-party application registry, a third-party application that is associated with the target third-party application object and starting the third-party application.

6. The method according to claim 5, wherein the selection instruction is a second speech signal;

and correspondingly,

the selecting a corresponding target third-party application object according to the received selection instruction specifically comprises;

converting the second speech signal to obtain a second text;

parsing the second text to acquire second target semantics; and

determining a corresponding target third-party application object according to the second target semantics.

7. The method according to claim 1, wherein before the acquiring, from a third-party application registry, a third-party application that is associated with the determined third-party application object and starting the third-party application, the method further comprises:

generating the third-party application registry.

8. The method according to claim 1, wherein after the starting the third-party application, the method further comprises:

sending a semantics command of the first target semantics to the third-party application, so that the third-party application invokes, according to the semantics command of the first target semantics, a third-party application interface API that matches the started third-party application, and the started third-party application acquires a required hardware or software resource by using the API.

9. The method according to claim 1, wherein after the starting the third-party application, the method further comprises:

receiving, by the third-party application, a third speech signal of the user;

converting the third speech signal to obtain a third text; and

invoking, according to a semantics command of the third text, a third-party application interface API that matches the started third-party application, so that the started third-party application acquires a required hardware or software resource by using the API.

10. The method according to claim 1, wherein the method is executed by an interaction device.

11. The method according to claim 1, wherein:

the parsing one or more first texts to obtain first target semantics is executed by a server; and

the determining, according to the acquired first target semantics, a third-party application object corresponding to the first target semantics and the acquiring, from a third-party application registry, a third-party application that is associated with the determined third-party application object and starting the third-party application are executed by an interaction device.

19. The interaction device according to claim 11, wherein:

after starting the third-party application, the starting unit sends a semantics command of the first target semantics to the third-party application, so that the third-party application invokes, according to the semantics command of the first target semantics, a third-party application interface API that matches the started third-party application.

26. The server according to claim 19, wherein:

after starting the third-party application, the starting unit sends a semantics command of the first target semantics to the third-party application, so that the third-party application invokes, according to the semantics command of the first target semantics, a third-party application interface API that matches the started third-party application.

12. An interaction device for speech recognition, wherein the interaction device comprises a parsing unit, a determining unit, a starting unit, and a storage unit, wherein:

the parsing unit is configured to parse one or more first texts to obtain first target semantics, wherein the first text is obtained through conversion from a first speech signal;

the determining unit is configured to determine, according to the first target semantics acquired by the parsing unit, a third-party application object corresponding to the first target semantics, wherein a third-party application that is associated with the third-party application object is a program that is not authorized to start by means of speech before delivery of a device;

the starting unit is configured to access the storage unit, acquire, from a third-party application registry stored by the storage unit, a third-party application that is associated with the third-party application object determined by the determining unit, and start the third-party application; and

the storage unit is configured to store the third-party application registry and the third-party application.

13. The interaction device according to claim 12, wherein the interaction device further comprises a speech recognizing unit, wherein:

the speech recognizing unit is configured to convert the first speech signals to obtain the one or more first texts and send the one or more first texts to the parsing unit.

14. The interaction device according to claim 12, wherein the interaction device further comprises a speech recognizing unit, wherein:

the speech recognizing unit is configured to convert the first speech signals to obtain at least two first texts and send the at least two first texts to the parsing unit; and

the parsing unit is specifically configured to score semantics of each first text of the at least two first texts according to a predetermined scoring rule and according to naturalness and coherence of the semantics of the at least two first texts, wherein a higher score represents better naturalness and coherence of the semantics; and

acquire, from the semantics of the at least two first texts, semantics with a highest score and of the first text as the first target semantics.

15. The interaction device according to claim 14, wherein the parsing unit is specifically configured to:

acquire, from the semantics of the at least two first texts, semantics, with a highest score, of the first text and exceeds a preset threshold as the first target semantics.

16. The interaction device according to claim 12, wherein the interaction device further comprises a sending unit and a receiving unit, wherein:

the determining unit is specifically configured to determine, according to the first target semantics, all third-party application objects corresponding to the first target semantics, and if the number of the third-party application objects is greater than or equal to two, acquire information about all the third-party application objects that are associated with the first target semantics;

the sending unit is configured to send prompt information, wherein the prompt information comprises the information about the third-party application objects and is used to prompt a user to further send, according to the information about the third-party application objects, a selection instruction, so that the determining unit selects a target third-party application object from all the third-party application objects that are associated with the first target semantics;

the receiving unit is configured to receive the selection instruction sent by the user; and

the determining unit is specifically configured to select a corresponding target third-party application object according to the received selection instruction; and

the starting unit acquires, from the third-party application registry stored by the storage unit, a third-party application that is associated with the target third-party application object and starts the third-party application.

17. The interaction device according to claim 16, wherein:

the speech recognizing unit is further configured to, when the selection instruction is a second speech signal, convert the second speech signal to obtain a second text and send the second text to the parsing unit; and

the parsing unit is further configured to parse the second text to acquire second target semantics; and

the determining unit is further configured to determine a corresponding target third-party application object according to the second target semantics.

18. The interaction device according to claim 12, wherein the interaction device further comprises a generating unit, wherein:

before the starting unit acquires, from the third-party application registry, the third-party application that is associated with the determined third-party application object and starts the third-party application, the generating unit generates the third-party application registry and sends the third-party application registry to the storage unit.

20. A server for speech recognition, comprising a parsing unit, a determining unit, a starting unit, and a storage unit, wherein:

the parsing unit is configured to parse one or more first texts to obtain first target semantics, wherein the first text is obtained through conversion from a first speech signal;

the determining unit is configured to determine, according to the first target semantics acquired by the parsing unit, a third-party application object corresponding to the first target semantics, wherein a third-party application that is associated with the third-party application object is a program that is not authorized to start by means of speech before delivery of a device;

the starting unit is configured to acquire, from a third-party application registry stored by the storage unit, a third-party application that is associated with the third-party application object determined by the determining unit and start the third-party application; and

the storage unit is configured to store the third-party application registry and the third-party application.

21. The server according to claim 20, wherein the parsing unit is specifically configured to:

score, according to a predetermined scoring rule and according to naturalness and coherence of semantics of at least two of the first texts that are obtained through conversion from the first speech signal, semantics of each first text of the at least two first texts, wherein a higher score represents better naturalness and coherence of the semantics; and

acquire, from the semantics of the at least two first texts, semantics with a highest score and of the first text as the first target semantics.

22. The server according to claim 21, wherein the parsing unit is specifically configured to:

acquire, from the semantics of the at least two first texts, semantics, with a highest score, of the first text and exceeds a preset threshold as the first target semantics.

23. The server according to claim 20, wherein the server further comprises a sending unit and a receiving unit, wherein:

the determining unit is specifically configured to determine, according to the first target semantics, all third-party application objects corresponding to the first target semantics, and if the number of the third-party application objects is greater than or equal to two, acquire information about all the third-party application objects that are associated with the first target semantics;

the sending unit is configured to send prompt information, wherein the prompt information comprises the information about the third-party application objects and is used to prompt a user to further send, according to the information about the third-party application objects, a selection instruction, so that the determining unit selects a target third-party application object from all the third-party application objects that are associated with the first target semantics;

the receiving unit is configured to receive the selection instruction sent by the user; and

the determining unit is specifically configured to select a corresponding target third-party application object according to the received selection instruction; and

the starting unit acquires, from the third-party application registry stored by the storage unit, a third-party application that is associated with the target third-party application object and starts the third-party application.

24. The server according to claim 23, wherein:

the parsing unit is further configured to, when the selection instruction is a second speech signal, parse the second text to acquire second target semantics, wherein the second text is obtained through conversion from the second speech signals; and

the determining unit is specifically configured to determine a corresponding target third-party application object according to the second target semantics.

25. The server according to claim 20, wherein the server further comprises a generating unit, wherein:

before the starting unit acquires, from the third-party application registry, the third-party application that is associated with the determined third-party application object and starts the third-party application, the generating unit generates the third-party application registry and sends the third-party application registry to the storage unit.

27. The server according to claim 20, wherein:

after the starting unit starts the third-party application, the third-party application converts the third speech signal to obtain a third text, and invokes, according to a semantics command of the third text, a third-party application interface API that matches the started third-party application, so that the started third-party application acquires a required hardware or software resource by using the API.

28. The server according to claim 20, wherein the receiving unit is further configured to receive, before the first text is parsed to obtain the first target semantics, the first text sent by an interaction device.

29. The server according to claim 20, wherein:

the sending unit is further configured to send a response feedback to the interaction device after the third-party application is started.