METHOD AND APPARATUS FOR PERFORMING AND CONTROLLING SPEECH RECOGNITION AND ENROLLMENT

• US 20120078637A1
• Filed: 12/09/2011
• Published: 03/29/2012
• Est. Priority Date: 06/12/2009
• 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: Zhu, Ning, Yang, Weiwei
• Granted Patent: US 8,909,533 B2
• Time in Patent Office: Days
• US Class Current: 704/270.1
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: H04L 65/1026 at the edge
• CPC Class Codes: H04L 65/103 in the network
• CPC Class Codes: H04L 65/1043 Gateway controllers, e.g. media gateway control protocol [MGCP] controllers

Abstract

A method and an apparatus for performing and controlling speech recognition and enrolment are provided. The method for performing speech recognition and enrolment includes: receiving a Speech Enrolment Start Request and a Speech Recognition Request sent from a media gateway controller (MGC); performing speech recognition and enrolment according to the Speech Enrolment Start Request and the Speech Recognition Request, and obtaining a recognition and enrolment result; and feeding back the recognition and enrolment result to the MGC.

Claims

1. A method for performing speech recognition and enrolment, comprising:

receiving a Speech Enrolment Start Request and a Speech Recognition Request from a media gateway controller (MGC);

performing speech recognition and enrolment according to the Speech Enrolment Start Request and the Speech Recognition Request, and obtaining a recognition and enrolment result; and

feeding back the recognition and enrolment result to the MGC.

2. The method for performing speech recognition and enrolment according to claim 1, wherein receiving a Speech Enrolment Start Request and a Speech Recognition Request from an MGC comprises:

receiving the Speech Enrolment Start Request from the MGC; and

after establishing an enrolment session according to the Speech Enrolment Start Request, receiving the Speech Recognition Request from the MGC.

3. The method for performing speech recognition and enrolment according to claim 1, wherein performing speech recognition and enrolment according to the Speech Enrolment Start Request and the Speech Recognition Request comprises:

initializing speech recognition and enrolment according to information in the Speech Enrolment Start Request; and

performing speech recognition and enrolment according to information in the Speech Recognition Request.

4. The method for performing speech recognition and enrolment according to claim 1, further comprising:

receiving an Enrolment Rollback Request from the MGC, and according to the Enrolment Rollback Request, giving up recording speech signals currently sent from a user equipment (UE);

receiving an End Phrase Enrolment Request from the MGC, and according to the End Phrase Enrolment Request, starting a new process of speech recognition and enrolment or aborting the current process of speech recognition and enrolment;

receiving a Modify Phrase Request from the MGC, and according to the Modify Phrase Request, modifying phrase parameter information corresponding to phrases that need to be modified; and

receiving a Delete Phrase Request from the MGC, and according to the Delete Phrase Request, deleting phrase parameter information corresponding to phrases that need to be deleted.

5. The method for performing speech recognition and enrolment according to claim 1, wherein the Speech Enrolment Start Request and the Speech Recognition Request are received through an H.248 signal.

6. The method for performing speech recognition and enrolment according to claim 5, wherein the Speech Enrolment Start Request is received through an extended signal enrolment start (enrst) in a package or an extended package of an H.248 protocol;

or the Speech Recognition Request is received through an extended protocol header field or a speech recognition signal of the H.248 protocol.

7. The method for performing speech recognition and enrolment according to claim 1, wherein the recognition and enrolment result comprises an enrolment failure or an enrolment success.

8. A method for controlling speech recognition and enrolment, comprising:

sending a Speech Enrolment Start Request and a Speech Recognition Request to a media gateway (MG); and

receiving a recognition and enrolment result fed back by the MG after the MG performs speech recognition and enrolment according to the Speech Enrolment Start Request and the Speech Recognition Request.

9. The method for controlling speech recognition and enrolment according to claim 8, wherein sending a Speech Enrolment Start Request and a Speech Recognition Request to an MG comprises:

sending the Speech Enrolment Start Request to the MG; and

sending the Speech Recognition Request to the MG after the MG establishes an enrolment session according to the Speech Enrolment Start Request.

10. The method for controlling speech recognition and enrolment according to claim 9, further comprising:

sending an Enrolment Rollback Request to the MG, so that according to the Enrolment Rollback Request, the MG gives up recording speech signals currently sent from a user equipment (UE);

sending an End Phrase Enrolment Request to the MG, so that according to the End Phrase Enrolment Request, the MG starts a new process of recognition and enrolment or aborts the current process of recognition and enrolment;

sending a Modify Phrase Request to the MG, so that according to the Modify Phrase Request, the MG modifies phrase parameter information corresponding to phrases that need to be modified; and

sending a Delete Phrase Request to the MG so that according to the Delete Phrase Request, the MG deletes phrase parameter information corresponding to phrases that need to be deleted.

11. The method for controlling speech recognition and enrolment according to claim 8, wherein the Speech Enrolment Start Request and the Speech Recognition Request are sent through an H.248 signal.

12. The method for controlling speech recognition and enrolment according to claim 11, wherein the Speech Enrolment Start Request is sent through an extended signal enrolment start (enrst) in a package or an extended package of an H.248 protocol;

or the Speech Recognition Request is sent through an extended protocol header field or a speech recognition signal of the H.248 protocol.

13. The method for controlling speech recognition and enrolment according to claim 8, wherein the recognition and enrolment result comprises an enrolment failure or an enrolment success.

14. A media gateway (MG), comprising:

a receiving module, configured to receive a Speech Enrolment Start Request and a Speech Recognition Request from a media gateway controller (MGC);

a speech recognizing and enrolling module, configured to perform speech recognition and enrolment according to the Speech Enrolment Start Request and the Speech Recognition Request received by the receiving module, and obtain a recognition and enrolment result; and

a feedback module, configured to feed back the recognition and enrolment result obtained by the speech recognizing and enrolling module to the MGC.

15. The MG according to claim 14, wherein the speech recognizing and enrolling module comprises:

an initializing unit, configured to initialize speech recognition and enrolment according to information carried in the Speech Enrolment Start Request; and

a recognizing and enrolling unit, configured to perform speech recognition and enrolment according to information carried in the Speech Recognition Request.

16. The MG according to claim 14, further comprising:

an enrolment rollback module, configured to receive an Enrolment Rollback Request from the MGC, and according to the Enrolment Rollback Request, give up recording speech signals currently from a user equipment (UE);

a phrase enrolment ending module, configured to receive an End Phrase Enrolment Request from the MGC, and according to the End Phrase Enrolment Request, start a new process of recognition and enrolment or abort the current process of recognition and enrolment;

a phrase modifying module, configured to receive a Modify Phrase Request from the MGC, and according to the Modify Phrase Request, modify phrase parameter information corresponding to phrases that need to be modified.

17. The MG according to claim 14, further comprising:

a phrase deleting module, configured to receive a Delete Phrase Request from the MGC, and according to the Delete Phrase Request, delete phrase parameter information corresponding to phrases that need to be deleted.

18. A media gateway controller (MGC), comprising:

a sending module, configured to send a Speech Enrolment Start Request and a Speech Recognition Request to a media gateway (MG); and

a result receiving module, configured to receive a recognition and enrolment result fed back by the MG after the MG performs speech recognition and enrolment according to the Speech Enrolment Start Request and the Speech Recognition Request.

19. The MGC according to claim 18, wherein the sending module is further configured to send an Enrolment Rollback Request to the MG, so that according to the Enrolment Rollback Request, the MG gives up recording speech signals currently from a user equipment (UE);

send an End Phrase Enrolment Request to the MG, so that according to the End Phrase Enrolment Request, the MG starts a new process of recognition and enrolment or aborts the current process of recognition and enrolment;

send a Modify Phrase Request to the MG, so that according to the Modify Phrase Request, the MG modifies phrase parameter information corresponding to phrases that need to be modified; and

/or send a Delete Phrase Request to the MG, so that according to the Delete Phrase Request, the MG deletes phrase parameter information corresponding to phrases that need to be deleted.

20. A system for speech recognition and enrolment, comprising a media gateway controller (MGC) and a media gateway (MG), wherein:

the MGC is configured to send a Speech Enrolment Start Request and a Speech Recognition Request to the MG; and

receive a recognition and enrolment result fed back by the MG after the MG performs speech recognition and enrolment according to information carried in the Speech Enrolment Start Request and the Speech Recognition Request; and

the MG is configured to receive the Speech Enrolment Start Request and the Speech Recognition Request from the MGC;

perform speech recognition and enrolment according to the information carried in the Speech Enrolment Start Request and the Speech Recognition Request, and obtain the recognition and enrolment result; and

feed back the recognition and enrolment result to the MGC.