Speech recognition system which turns its voice response on for confirmation when it has been turned off without confirmation

• US 5,983,179 A
• Filed: 06/26/1997
• Issued: 11/09/1999
• Est. Priority Date: 11/13/1992
• Status: Expired due to Term

• Current Assignee: Nuance Communications, Inc. (Microsoft Corporation)
• Original Assignee: Dragon Systems, Inc. (Microsoft Corporation)
• Inventors: Gould, Joel M.
• Primary Examiner(s): Dorvil, Richemond
• Application Number:
  US08/882,918
• Time in Patent Office: 866 Days
• Field of Search: 704/231, 704/251, 704/270, 704/273, 704/272, 704/275, 704/255, 704/246, 704/247, 704/250
• US Class Current: 704/270.1
• CPC Class Codes: G09B 19/04 Speaking with audible presentation of the material to be studied G09B5/04
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: G10L 15/18 using natural language modelling
• CPC Class Codes: G10L 15/1815 Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 15/285 Memory allocation or algorithm optimisation to reduce hardware requirements
• CPC Class Codes: G10L 2015/0631 Creating reference templates; Clustering
• CPC Class Codes: G10L 2015/0638 Interactive procedures
• CPC Class Codes: G10L 2015/228 of application context

Abstract

A speech recognition system includes a speech-response capability for responding to sounds which appear to match models of spoken words by performing functions associated with such words. This speech response can be turned on or off. If the system detects both that speech response is off and that no indication is stored that the off state has been confirmed by a user, it performs a confirmation process. This prompts the user to utter a phrase confirming whether or not speech response is to be off; turns speech response on so it can respond to the user'"'"'s confirmation utterance; determines whether to leave speech response on or off in response to the word which the speech response selects as corresponding to the user'"'"'s confirmation utterance; and responds to such a determination that speech response is to be turned off, both by turning it off and by storing an indication that the off state has been confirmed by the user. In some embodiments, the system can execute user written programs which can include instructions for turning speech response on or off. In this case, the system can delay the start of a confirmation process until after the execution of such a program, to allow such programs to temporarily turn off speech response without requiring user confirmation. In some embodiments, the user can select whether to deactivate the feature which causes a confirmation processes to occur whenever speech response has been turned off without confirmation.

Claims

1. A computerized speech recognition system comprising:

speech-response means including;

means for receiving acoustics signals representing sounds received by a microphone;

recognition means for pattern matching said acoustic signals against models of vocabulary words, to select, based on said pattern matching, which vocabulary words, if any, most probably correspond to those signals; and

means for responding to a selection of a given vocabulary word as most probably corresponding to a given acoustic signal by performing a function associated with that vocabulary word;

on-off means for selectively setting the speech response means to an on or off state in which said speech response means, respectively, either can or cannot, respond to sounds by performing said associated functions;

means for storing confirmation information which indicates whether or not an off state of said speech response means has been confirmed by the user;

confirmation means for performing a confirmation process including;

first means for prompting the user to utter a phrase of one or more words confirming whether or not the user wants said speech response means to be in the off state;

second means for causing said on-off means to set said speech response means to an on state so said speech response means can respond to an utterance the user makes in response to said prompt;

third means for determining whether to place said speech response means in said on or off state in response to the vocabulary word selected by the recognition means as corresponding to an acoustic signal received after said prompt;

fourth means for responding to such a determination to place said speech response means off, by causing said on-off means to set said speech response means to said off state and by storing confirmation information indicating that the off state has been confirmed by the user; and

means, responsive to the combination of said speech response means being in said off state and said confirmation information indicating that the off state has not been confirmed by the user, for initiating said confirmation process in which said first, second, third, and forth means each perform their respective function.

2. A computerized speech recognition system as in claim 1:

further including means for executing a user written sequence of programmed instructions for controlling the operation of said speech recognition program;

wherein said on-off means includes means for setting said speech response means to either said on or off state in response to different instructions which can be included in said sequence of programmed instructions; and

said confirmation means includes means for delaying the initiation of the confirmation process until after said system is done executing such a sequence of programmed instructions.

3. A computerized speech recognition system as in claim 1 further including means for enabling a user to selectively enable or disable the operation of said confirmation means.

4. A computer program recorded in machine readable memory for performing speech recognition, said program comprising:

speech-response instructions including;

instructions for receiving acoustics signals representing sounds received by a microphone;

recognition instructions for pattern matching said acoustic signals against models of vocabulary words, to select, based on said pattern matching, which vocabulary words, if any, most probably correspond to those signals; and

instructions for responding to a selection of a given vocabulary word as most probably corresponding to a given acoustic signal by performing a function associated with that vocabulary word;

on-off instructions for enabling the system'"'"'s speech response to selectively be set to either an on or off state in which said speech response instructions, respectively, either can or cannot, respond to sounds by performing said associated functions;

instructions for storing confirmation information which indicates whether or not an off state of said speech response has been confirmed by the user;

confirmation instructions for performing a confirmation process including;

first instructions for prompting the user to utter a phrase of one or more words confirming whether or not the user wants said speech response to be in the off state;

second instructions for causing said on-off instructions to set said speech response to an on state so said speech response instructions can respond to an utterance the user makes in response to said prompt;

third instructions for determining whether to place said speech response in said on or off state in response to the vocabulary word selected by the recognition instructions as corresponding to an acoustic signal received after said prompt;

fourth instructions for responding to such a determination to place said speech response off, by causing said on-off instructions to set said speech response to said off state and by storing confirmation information indicating that the off state has been confirmed by the user; and

instructions, responsive to the combination of said speech response being in said off state and said confirmation information indicating that the off state has not been confirmed by the user, for initiating said confirmation process in which said first, second, third, and forth instructions each perform their respective function.

5. A computer program as in claim 4:

further including instructions for executing a user written program for controlling the operation of said speech recognition program;

wherein said on-off instructions includes instructions for setting said speech response to either said on or off state in response to commands which can be included in said user written program; and

said confirmation instructions includes instructions for delaying the initiation of the confirmation process until after said system is done executing said user written program.

6. A computer program as in claim 4 further including instructions for enabling a user to selectively enable or disable the operation of said confirmation instructions.

7. A computerized speech recognition method comprising the steps of:

peforming speech-response, including the steps of;

receiving acoustics signals representing sounds received by a microphone;

pattern matching said acoustic signals against models of vocabulary words, to select, based on said pattern matching, which vocabulary words, if any, most probably correspond to those signals; and

responding to a selection of a given vocabulary word as most probably corresponding to a given acoustic signal by performing a function associated with that vocabulary word;

enabling the system'"'"'s speech response to selectively be set to either an on or off state in which said system, respectively, either can or cannot respond to sounds by performing said associated functions;

storing confirmation information which indicates whether or not an off state of said speech response has been confirmed by the user;

responding to the combination of said speech response being in said off state and said confirmation information indicating that the off state has not been confirmed by the user, by performing a confirmation process including the steps of;

prompting the user to utter a phrase of one or more words confirming whether or not the user wants said speech response to be in the off state;

setting said speech response to an on state so said speech response can respond to an utterance the user makes in response to said prompt;

determining whether to place said speech response in said on or off state in response to the vocabulary word selected by said pattern matching as corresponding to an acoustic signal received after said prompt; and

responding to such a determination to place said speech response off, by setting said speech response to said off state and by storing confirmation information indicating that the off state has been confirmed by the user.

8. A computerized speech recognition method as in claim 7:

further including the step of executing a user written sequence of programmed instructions for controlling the operation of said speech recognition program;

wherein said step of enabling speech response to be set includes enabling said speech response to be set to either said on or off states in response to instructions in said sequence of programmed instructions; and

said step which performs said confirmation process delays the initiation of the confirmation process until after said system is done executing such a sequence of programmed instructions.

9. A computerized speech recognition method as in claim 7 further including the steps of enabling a user to selectively enable or disable the performance of said confirmation process.