Method for operating a speech recognition system

• US 8,781,826 B2
• Filed: 10/24/2003
• Issued: 07/15/2014
• Est. Priority Date: 11/02/2002
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Kooiman, Albert
• Primary Examiner(s): Shah, Paras D
• Application Number:
  US10/532,919
• Publication Number:
  US 20060200345A1
• Time in Patent Office: 3,917 Days
• Field of Search: 704/205, 704/226, 704/228, 704/233, 704/231, 704/246, 455/101, 455/135, 455/226.1, 455/277.2, 455/423, 455/436, 455/437, 455/522, 455/525, 455/561, 455/562.1, 455/63.1, 455/67.11, 455/67.13, 455/69, 375/130, 375/150, 375/224, 375/260, 375/267, 375/340, 375/346, 375/347, 375/E1.004, 370/208, 370/210, 370/235, 370/252, 370/328, 370/329, 370/331, 370/332, 370/335, 370/338, 370/342, 370/389
• US Class Current: 704/233
• CPC Class Codes: G10L 15/20 Speech recognition techniques specially adapted for robustness in adverse environments, e.g. in noise, of stress induced speech G10L21/02 takes precedence
• CPC Class Codes: G10L 2015/225 Feedback of the input speech
• CPC Class Codes: G10L 2015/226 using non-speech characteristics

Abstract

A method for operating a speech recognition system is described in which a speech signal (S₁) of a user is detected and analyzed so as to recognize speech information contained in the speech signal (S₁). The speech recognition system determines a reception quality value (S_Q) or a noise value which represents a current reception quality. The speech recognition system is switched over to a mode of operation which is less sensitive to noise and/or outputs an alert signal (S_W) to the user when the reception quality value (S_Q) drops below a given reception quality threshold or when the noise value exceeds a noise threshold. An appropriate speech recognition system is also described.

Claims

1. A method for operating a speech recognition system comprising acts of:

detecting a speech signal of a user;

analyzing the speech signal to recognize speech information contained in the speech signal;

determining a reception quality value or a noise value which represents a current reception quality; and

switching the speech recognition system over to a mode of operation which is less sensitive to noise in response to determining that the noise value has exceeded a noise threshold or that the reception quality value has dropped below a reception quality threshold, or both; and

automatically resetting the speech recognition system to a previous mode of operation in response to determining that the reception quality value has exceeded the reception quality threshold or that the noise value has dropped below the noise threshold.

2. The method of claim 1, wherein the act of switching the speech recognition system over to the mode of operation which is less sensitive to noise comprises deactivating a capability of the speech recognition system to operate in a barge-in mode of operation so that the speech recognition system does not react to speech input of the user during output of a prompt.

3. The method of claim 1, wherein the reception quality value or the noise value is determined with a voice activity detector.

4. The method of claim 1, wherein the reception quality value or the noise value is determined on the basis of a background signal which is received prior to a beginning of an utterance, or in a speech pause of the user, or both.

5. The method of claim 1, further comprising an act of:

in response to determining that the reception quality value has dropped below the reception quality threshold or that the noise value has exceeded the noise threshold, applying a reception corruption indication signal to a dialog control device.

6. The method of claim 5, further comprising:

in response to the dialog control device receiving the reception corruption indication signal, using the dialog control device to initiate an output of a prompt indicating that the reception conditions are poor.

7. The method of claim 1, further comprising acts of:

analyzing an incoming signal for a type of disturbance causing the reception quality value to be below the reception quality threshold or the noise value to be above the noise threshold, and

in response to identifying the type of disturbance in the incoming signal, outputting a prompt to the user.

8. A speech recognition system, comprising:

at least one processor configured to detect a speech signal of a user;

to analyze the detected speech signal to recognize speech information contained in the speech signal;

to determine a reception quality value or a noise value, representing a current reception quality;

to compare the reception quality value with a reception quality threshold or to compare the noise value with a noise threshold;

to switch the speech recognition system over to a mode of operation which is less sensitive to noise in response to determining that the reception quality value has dropped below the reception quality threshold or that the noise value has exceeded the noise threshold, or both, and to automatically reset the speech recognition system to a previous mode of operation in response to determining that the reception quality value has exceeded the reception quality threshold or that the noise value has dropped below the noise threshold.

10. The system of claim 8, wherein the at least one processor is further configured to detect the speech signal by detecting voice activity of the user.

11. The system of claim 8, wherein the at least one processor is further configured to switch the speech recognition system over to the mode of operation which is less sensitive to noise by disabling a barge-in switching functionality so that the speech recognition system does not react to speech input of the user during output of a prompt.

12. The system of claim 8, wherein the at least one processor is further configured to control dialog based on the speech information contained in the speech signal.

9. A non-transitory, computer readable medium encoded with instructions which, when executed by at least one processor, cause the at least one processor to carry out a method for operating a speech recognition system, the method comprising acts of:

detecting a speech signal of a user;

analyzing the speech signal to recognize speech information contained in the speech signal;

determining a reception quality value or a noise value which represents a current reception quality;

switching the speech recognition system over to a mode of operation which is less sensitive to noise in response to determining that the noise value has exceeded a noise threshold or that the reception quality value has dropped below a reception quality threshold, or both; and

automatically resetting the speech recognition system to a previous mode of operation in response to determining that the reception quality value has exceeded the reception quality threshold or that the noise value has dropped below the noise threshold.

13. The computer readable medium of claim 9, wherein the act of switching the speech recognition system over to the mode of operation which is less sensitive to noise comprises deactivating a capability of the speech recognition system to operate in a barge-in mode of operation so that the speech recognition system does not react to speech input of the user during output of a prompt.

14. A method for operating a speech recognition system, the method comprising:

detecting a speech signal (SI) of a user;

analyzing the speech signal to recognize speech information contained in the speech signal (SI);

determining a reception quality value (SQ) or a noise value which represents a current reception quality;

switching the speech recognition system over to a mode of operation, which is less sensitive to noise when the noise value exceeds a noise threshold or outputting an alert signal (SW) to the user when the reception quality value (SQ) drops below a given reception quality threshold, or both; and

automatically resetting the speech recognition system to a previous mode of operation when the reception quality value (SQ) exceeds the reception quality threshold or when the noise value drops below the noise threshold.

15. The method of claim 14, wherein the act of switching the speech recognition system over to the mode of operation which is less sensitive to noise comprises deactivating a capability of the speech recognition system to operate in a barge-in mode of operation so that the speech recognition system does not react to speech input of the user during output of a prompt.

16. The method of claim 14, wherein the reception quality value or the noise value is determined with a voice activity detector.

17. The method of claim 14, wherein the reception quality value or the noise value is determined on the basis of a background signal which is received prior to a beginning of an utterance, or in a speech pause of the user, or both.

18. The method of claim 14, further comprising an act of:

in response to determining that the reception quality value has dropped below the reception quality threshold or that the noise value has exceeded the noise threshold, applying a reception corruption indication signal to a dialog control device.

19. The method of claim 17, further comprising:

in response to the dialog control device receiving the reception corruption indication signal, using the dialog control device to initiate an output of a prompt indicating that the reception conditions are poor.

20. The method of claim 14, further comprising acts of:

analyzing an incoming signal for a type of disturbance causing the reception quality value to be below the reception quality threshold or the noise value to be above the noise threshold, and

in response to identifying the type of disturbance in the incoming signal, outputting a prompt to the user.