Distributed speech processing

• US 20040107107A1
• Filed: 12/03/2002
• Published: 06/03/2004
• Est. Priority Date: 12/03/2002
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: SpeechWorks International, Inc. (Microsoft Corporation)
• Inventors: Corriveau, Francois, Lenir, Philip, Hunt, Andrew
• Granted Patent: US 7,571,100 B2
• Time in Patent Office: Days
• US Class Current: 704/270.1
• CPC Class Codes: G06F 21/32 using biometric data, e.g. fingerprints, iris scans or voiceprints
• CPC Class Codes: G10L 15/30 Distributed recognition, e.g. in client-server systems, for mobile phones or network applications
• CPC Class Codes: G10L 17/22 Interactive procedures; Man-machine interfaces

Abstract

Processing a speech utterance by communicating between a local computer and a remote computer using a hyper text communication session. The local computer sends a recording of a speech utterance to the remote computer in the session, and receives a result from the remote computer, the result based on a processing of the recording at the remote computer.

Claims

1. A method for processing a speech utterance, comprising:

communicating between a local computer and a remote computer using a hyper text communication session, including

sending a recording of a speech utterance from the local computer to the remote computer in the session, and

receiving a result from the remote computer, the result based on a processing of the recording at the remote computer.

2. The method of claim 1 in which the result characterizes a similarity of voice characteristics.

3. The method of claim 1 in which the result indicates that a speaker who made the speech utterance is a known person.

4. The method of claim 1 in which the result indicates a match of stored voice characteristics of a speaker.

5. The method of claim 1, further comprising prompting the user.

6. The method of claim 5, further comprising recording a speech utterance from the user using the local computer.

7. The method of claim 5, further comprising using the local computer to receive a script that includes an instruction that instructs the local computer to prompt the user.

8. The method of claim 1, further comprising using the local computer to receive a script that includes a universal resource locator of an application program that is run by the remote computer to process the recording.

9. The method of claim 8 in which the script includes an instruction that instructs the local computer to perform a task based on the result received from the remote computer.

10. The method of claim 8 in which the script includes extensible markup language tags.

11. The method of claim 10 in which the script includes voice extensible markup language tags.

12. The method of claim 1, further comprising processing the recording to determine a similarity of voices.

13. The method of claim 1, further comprising processing the recording to indicate a likelihood that the speaker is a known person.

14. The method of claim 1, further comprising processing the recording to indicate a match of stored voice characteristics.

15. The method of claim 14, further comprising using the local computer to prompt a user to make the speech utterance.

16. The method of claim 1 in which the hyper text communication session follows a hyper text transfer protocol.

17. The method of claim 16 in which the hyper text transfer protocol is an HTTP protocol defined by World Wide Web Consortium.

18. A method comprising:

receiving a dialog file at a web browser;

controlling a speech dialog using the received dialog file;

receiving a speech utterance from a user as part of the speech dialog;

encoding the speech utterance to generate an encoded speech utterance;

sending a request from the web browser to a web server, the request containing the encoded speech utterance; and

receiving a response from the web server, the response containing a result based on a processing of the encoded speech utterance.

19. The method of claim 18 in which the request contains an identifier to an application used to process the speech utterance.

20. The method of claim 18 in which receiving the dialog file comprises receiving the dialog file from the web server.

21. The method of claim 18 in which the dialog file comprises a VoiceXML document.

22. The method of claim 18 in which the encoded speech utterance comprises an MIME-encoded message.

23. The method of claim 18 in which the request is sent from the web browser to the web server according to Hypertext Transfer Protocol.

24. The method of claim 23 in which the response is sent from the web server to the web browser according to the Hypertext Transfer Protocol.

25. The method of claim 18 in which the application comprises a speech recognizer.

26. The method of claim 18 wherein the application is a speech recognizer and the result characterizes a similarity of voice characteristics.

27. A method comprising:

sending a dialog file from a server to a client, the dialog file containing statements for processing by the client to control a speech dialog;

receiving at a server a request from the client in response to the client processing one of the statements, the request containing an encoded speech utterance;

processing the encoded speech utterance by using the server; and

sending a response from the server to the client, the response containing a result based on processing of the encoded speech utterance.

28. The method of claim 27 in which the dialog file comprises a VoiceXML document.

29. The method of claim 27 in which the encoded speech utterance comprises an MIME-encoded message.

30. The method of claim 27 in which the request is sent from the client to the server according to Hypertext Transfer Protocol.

31. The method of claim 30 in which the response is sent from the server to the client according to the Hypertext Transfer Protocol.

32. The method of claim 27 in which processing the encoded speech utterance includes applying a speech recognition application to process the encoded speech.

33. The method of claim 27 in which processing the encoded speech utterance includes applying a speaker verification application to process the encoded speech.

34. The method of claim 27 in which the result characterizes a similarity of voice characteristics.

35. A method comprising:

receiving a speech utterance from a user;

encoding the speech utterance to generate an encoded speech utterance;

sending a request through a network to a server in a hyper text communication session, the request containing the encoded speech utterance and an identifier to an application used to process the speech utterance; and

receiving a response from the server that contains a result based on a processing of the encoded speech utterance.

36. The method of claim 35 in which the request comprises an HTTP POST request.

37. The method of claim 35 in which the hyper text communication session comprises a series of related HTTP requests and responses.

38. An apparatus comprising:

means for communicating with a remote computer using a hyper text communication session;

means for sending a recording of a speech utterance to the remote computer in the session; and

means for receiving a result from the remote computer, the result based on a processing of the recording at the remote computer.

39. The apparatus of claim 38, further comprising means for receiving the speech utterance from a user and converting the speech utterance into a recording.

40. The apparatus of claim 39 wherein the means for sending the recording comprises means for converting the recording to a Multipurpose Internet Mail Extension (MIME)-encoded message.

41. The apparatus of claim 40 wherein the means for sending the recording comprises means for sending the MIME-encoded message using a Hypertext Transfer Protocol (HTTP) POST command.

42. The apparatus of claim 38, further comprising means for receiving a script that includes an instruction that instructs the apparatus to prompt a user.

43. The apparatus of claim 42 wherein the script includes extensible markup language tags.

44. The apparatus of claim 43, further comprising means for interpreting the extensible markup language tags.

45. Software stored on computer-readable media for causing a computer system to perform functions comprising:

communicating between a local computer and a remote computer using a hyper text communication session, including

sending a recording of a speech utterance from the local computer to the remote computer in the session, and

receiving a result from the remote computer, the result based on a processing of the recording at the remote computer.

46. Software stored on computer-readable media for causing a computer system to perform functions comprising:

receiving a dialog file at a web browser;

controlling a speech dialog using the received dialog file;

receiving a speech utterance from a user as part of the speech dialog;

encoding the speech utterance to generate an encoded speech utterance;

sending a request from the web browser to a web server, the request containing the encoded speech utterance; and

receiving a response from the web server, the response containing a result based on a processing of the encoded speech utterance.

47. Software stored on computer-readable media for causing a computer system to perform functions comprising:

sending a dialog file from a server to a client, the dialog file containing statements for processing by the client to control a speech dialog;

receiving at a server a request from the client in response to the client processing one of the statements, the request containing an encoded speech utterance;

processing the encoded speech utterance by using the server; and

sending a response from the server to the client, the response containing a result based on processing of the encoded speech utterance.

48. Software stored on computer-readable media for causing a computer system to perform functions comprising:

receiving a speech utterance from a user;

encoding the speech utterance to generate an encoded speech utterance;

sending a request through a network to a server in a hyper text communication session, the request containing the encoded speech utterance and an identifier to an application used to process the speech utterance; and

receiving a response from the server that contains a result based on a processing of the encoded speech utterance.

49. An apparatus comprising:

an input port to receive a speech utterance from a user as part of a speech dialog; and

a web browser to receive a dialog file and control the speech dialog using the received dialog file, the web browser being configured to encode the speech utterance to generate an encoded speech utterance, to send a request containing the encoded speech utterance to a web server, and to receive a response containing a result based on a processing of the encoded speech utterance from the web server.

50. A server computer comprising:

a storage to store a dialog file containing statements for processing by a client to control a speech dialog;

an input/output port to send the dialog file to the client and to receive a request from the client in response to the client processing one of the statements, the request containing an encoded speech utterance; and

an application to process the encoded speech utterance and to send a response containing a result based on processing of the encoded speech utterance to the client.

51. A voice-enabled device comprising:

an input/output interface to receive a speech utterance from a user;

a voice-enabled application configured to encode the speech utterance to generate an encoded speech utterance and send a request through a network to a server in a hyper text communication session, the request containing the encoded speech utterance and an identifier to an application used to process the speech utterance, the voice-enabled application further configured to receive a response from the server that contains a result based on a processing of the encoded speech utterance and to perform a function based on the result.

52. A telephone call center comprising:

a call manager to receive a speech utterance of a user transmitted through a telephone network, the call manager configured to determine a telephone number dialed by the user to connect the user to the telephone call center, the call manager further configured determine a universal resource locator (URL) based on the telephone number; and

a computer to run a speech browser application that performs the functions of

retrieving a script based on the URL provided by the call manager,

encoding the speech utterance into an encoded speech utterance,

sending a request through a network to a server in a hyper text communication session, the request containing the encoded speech utterance and an identifier to an application used to process the speech utterance, and

receiving a response from the server that contains a result based on a processing of the encoded speech utterance.