Componentized voice server with selectable internal and external speech detectors

• US 7,925,510 B2
• Filed: 04/28/2004
• Issued: 04/12/2011
• Est. Priority Date: 04/28/2004
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Inventors: Creamer, Thomas E., Moore, Victor S., Nusbickel, Wendi L., Sliwa, James J., Dos Santos, Ricardo
• Primary Examiner(s): Wozniak; James S
• Assistant Examiner(s): Shah; Paras
• Application Number:
  US10/833,615
• Publication Number:
  US 20050246166A1
• Time in Patent Office: 2,540 Days
• Field of Search: 704/233, 704/270.1, 704/270, 704/275, 379/88.04
• US Class Current: 704/270.1
• CPC Class Codes: G10L 25/78 Detection of presence or absence of voice signals switching of direction of transmission by voice frequency in two-way loud-speaking telephone systems H04M9/10

Abstract

A method for detecting speech utterances within a telephone call can include the steps of initializing a componentized voice server having at least one software-based speech detection routine. At least one previously established parameter can be used to discern a speech detection methodology for handling an incoming call. The software-based speech detection routine can be set in accordance with a select one of the parameters. An indicator of particular one of the parameters can be conveyed to an external speech detection component so that the external speech detection component is set to detect speech for the call in accordance with the conveyed indication. The software-based speech detection routine and/or the external speech detection component can detect a speech utterance for the call. The voice server can perform at least one programmatic action responsive to the detecting of the speech utterance.

Claims

1. A method for detecting speech utterances within a telephone call comprising the steps of:

receiving a signal representing a telephone call received over a telephone network by a telephone gateway;

initializing a componentized voice server having an internal speech detection module with a plurality of software-based speech detection routines and a Pluggable, configurable external speech detection component operationally located remotely from the voice server, wherein the external speech detection component is implemented as an electronic module plugged into a piece of equipment coupled in a signal path between the telephone network and the voice server;

presenting through a user interface options for speech detection settings and receiving through the user interface user selections indicating speech detection parameters, wherein the speech detection parameters determine whether the internal speech detection module, the external speech detection component or both the internal speech detection module and the external speech detection component will be activated;

when the received speech detection parameters indicate that the external speech detection component will be activated;

sending a message from the voice server to the external speech detection component to activate said external speech detection component;

processing the received signal to detect a speech utterance within the signal using the activated external speech detection component;

sending a message from the external speech detection component to the voice server conveying results of detecting a speech utterance; and

performing with said voice server at least one programmatic action responsive to the detecting of the speech utterance, the programmatic action comprising recognizing speech in the detected speech utterance; and

when the received speech detection parameters indicate that both the internal speech detection module and the external speech detection component will be activated;

sending a message from the voice server to the external speech detection component to activate said external speech detection component;

processing the received signal using the activated external speech detection component;

sending a message from the external speech detection component to the voice server conveying results of an attempt to detect a speech utterance; and

performing with said voice server at least one programmatic action, the programmatic action comprising using the internal speech detection module conjunctively with the results of the attempt to detect the speech utterance in the external speech detection component to detect the speech utterance in the received signal; and

when the received speech detection parameters indicate that the internal speech detection module will be activated;

processing the received signal to detect a speech utterance within the signal using the internal speech detection module; and

performing with said voice server at least one programmatic action responsive to the detecting of the speech utterance.

2. The method of claim 1, wherein both of said internal speech detection module and said external speech detection component technique are utilized simultaneously.

3. The method of claim 1, wherein said external speech detection component is utilized and said internal speech detection module is not utilized.

4. The method of claim 1, wherein said external speech detection component performs hardware-based speech detection.

5. The method of claim 1, wherein said external speech detection component detects speech by detecting energy differences within a telephony channel.

6. The method of claim 1, further comprising the step of:

before said initializing step, receiving a user specified parameter; and

storing said user specified parameter in a data store communicatively linked to said voice server.

7. The method of claim 1, wherein sending the message from the voice server to the external speech detection component comprises sending parameters relating to speech detection, the parameters comprising one or more of a threshold value, a preferred detection algorithm, and a characterization of speech utterances to be detected.

8. The method of claim 1, wherein:

the speech detection parameters further determine which detection routine within the internal speech detection module will be activated if the internal speech detection module is selected;

when the received speech detection parameters indicate that both the internal speech detection module and the external speech detection component will be activated, the method further comprises activating the detection routine within the internal speech detection module determined by the received speech detection parameters; and

when the received speech detection parameters indicate that the internal speech detection module will be activated, the method further comprises activating the detection routine within the internal speech detection module determined by the received speech detection parameters.

9. The method of claim 1, wherein presenting through the user interface options for speech detection settings comprises presenting a list of the options for speech detection settings.

10. A method for detecting speech utterances within a telephone call, the method comprising:

initializing a voice server having an internal speech detection module with a plurality of software-based speech detection routines;

initializing a configurable external speech detection component operationally located remotely from the voice server to process a received call, wherein;

the external speech detection component is incorporated in a piece of equipment coupled in a signal path between a telephone network and the voice server,

the initializing the configurable external speech detection component is performed based on speech detection parameters established prior to receiving the call, and

the speech detection parameters identify speech detection processing to be performed on the telephone call;

when the speech detection parameters indicate processing external to the voice server, activating the external speech detection component and conveying to the external speech detection component an indication of a parameter of the speech detection parameters;

receiving the telephone call through a telephone gateway;

processing the telephone call in the external speech detection component to detect a speech utterance within the telephone call in accordance with the indicated parameter;

providing the results of the processing to the voice server; and

in response to the provided results, performing in the voice server at least one programmatic action responsive to the detecting of the speech utterance, the programmatic action comprising recognizing speech within the detected speech utterance,

wherein the parameters comprise a threshold value, a preferred detection algorithm, and a characterization of speech utterances to be detected.

11. The method of claim 10, wherein the method further comprises:

activating an internal speech detection module; and

processing the telephone call in the internal speech detection module.

12. The method of claim 11, wherein:

activating an internal speech detection module comprises activating the internal speech detection module to perform speech detection using a selected software-based speech detection routine of the plurality of software-based speech detection routines.

13. The method of claim 11, wherein:

the results of speech detection performed in the external speech detection component and the internal speech detection module are used conjunctively.

14. The method of claim 11, wherein:

the previously established speech detection parameters comprise parameters received through a user interface from each of a plurality of customers; and

the initializing is performed based on parameters associated with a customer making the call.

15. The method of claim 11, wherein:

initializing the voice server and initializing the configurable external speech detection component are performed in response to receiving the call.

16. A system for detecting speech utterances within a telephone call, the system comprising:

a telephone gateway adapted to be coupled to a telephone network to receive signals representative of telephone calls over the telephone network;

a voice server coupled to the telephone gateway, the voice server comprising computer storage media storing computer executable instructions;

a voice detection unit in a piece of equipment connected to couple a telephone call from the telephone network to the voice server,

wherein the computer executable instructions comprise instructions, when executed, for;

recognizing content of speech within a detected speech utterance;

receiving user input indicating a level of speech detection and determining based on the user input a parameter for processing to detect speech utterances;

sending a message to the voice detection unit to configure the voice detection unit in accordance with the indicated parameter and to configure a speech detection setting within the voice detection unit based on the indicated parameter;

receiving a result of voice detection from the voice detection unit; and

in response to the received results, performing at least one programmatic action.

17. The system of claim 16, wherein the piece of equipment comprises an interface slot and the voice detection unit is plugged into the interface slot.

18. The system of claim 17, wherein the piece of equipment comprises a router and the voice detection unit is plugged into the interface slot of the router.

19. The system of claim 17, wherein the piece of equipment comprises a telephony switch and the voice detection unit is plugged into the interface slot of the telephony switch.

20. The system of claim 17, wherein the piece of equipment comprises the telephone gateway and the voice detection unit is plugged into the interface slot of the telephone gateway.