Method and system for editing phrases during continuous speech recognition

• US 5,884,258 A
• Filed: 10/31/1996
• Issued: 03/16/1999
• Est. Priority Date: 10/31/1996
• Status: Expired due to Term

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Alleva, Fileno A, Rozak, Michael J.
• Primary Examiner(s): Hudspeth, David R.
• Assistant Examiner(s): SAX, ROBERT L
• Application Number:
  US08/741,691
• Time in Patent Office: 866 Days
• Field of Search: 704/251, 704/244, 704/246, 704/245, 704/231, 704/233, 704/234
• US Class Current: 704/251
• CPC Class Codes: G06F 3/16 Sound input; Sound output speech processing G10L
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/0631 Creating reference templates; Clustering
• CPC Class Codes: G10L 2015/221 Announcement of recognition results

Abstract

A method and system for editing words that have been misrecognized. The system allows a speaker to specify a number of alternative words to be displayed in a correction window by resizing the correction window. The system also displays the words in the correction window in alphabetical order. A preferred system eliminates the possibility, when a misrecognized word is respoken, that the respoken utterance will be again recognized as the same misrecognized word. The system, when operating with a word processor, allows the speaker to specify the amount of speech that is buffered before transferring to the word processor.

Claims

1. A method in a continuous-speech recognition computer system for correcting speech that is misrecognized;

the speech being recognized as a plurality of words, the method comprising;

receiving from a speaker a selection of a misrecognized phrase comprising less than all of the plurality of words, the misrecognized phrase having two or more words;

after receiving the selection of the misrecognized phrase, identifying one or more alternative phrases that may correspond to the misrecognized phrase;

displaying the identified alternative phrases;

receiving from the speaker a selection of a displayed phrase; and

replacing the misrecognized phrase with the selected phrase.

2. A method in a continuous-speech recognition computer system for correcting a spoken utterance that is misrecognized as a phrase having a plurality of words, the method comprising:

receiving from a speaker an indication to correct the misrecognized phrase;

identifying one or more alternative phrases that may correspond to the spoken utterance, wherein the identifying of alternative phrases does not identify any alternative phrase that differs from the misrecognized phrase by only one word;

displaying the identified alternative phrases;

receiving from the speaker a selection of a displayed phrase; and

replacing the misrecognized phrase with the selected phrase.

3. A method in a continuous-speech recognition computer system for correcting a spoken utterance that is misrecognized as a phrase having a plurality of words, the method comprising:

receiving from a speaker an indication to correct the misrecognized phrase;

identifying one or more alternative phrases that may correspond to the spoken utterance, wherein the identifying of alternative phrases does not identify any alternative phrase that can be corrected by selecting alternative words for each word of the misrecognized phrase;

displaying the identified alternative phrases;

receiving from the speaker a selection of a displayed phrase; and

replacing the misrecognized phrase with the selected phrase.

4. A computer-readable medium having instructions for causing a continuous-speech recognition computer system to correct speech that is misrecognized, the speech being recognized as a plurality of words, by:

displaying the plurality of words;

receiving from a speaker a selection of a misrecognized phrase comprising less than all of the plurality of words, the misrecognized phrase having two or more words;

after receiving the selection of the misrecognized phrase, identifying a plurality of alternative phrases that may correspond to the misrecognized phrase;

displaying the identified alternative phrases;

receiving from the speaker a selection of a displayed phrase; and

displaying the selected phrase in place of the misrecognized phrase.

7. The computer-readable medium of claim 4 wherein the identifying of alternative phrases includes inputting and recognizing a newly spoken utterance for the misrecognized phrase.

5. A computer-readable medium having instructions for causing a continuous-speech recognition computer system to correct a spoken utterance that is misrecognized as a phrase having a plurality of words, by:

displaying the misrecognized phrase;

receiving from a speaker an indication to correct the misrecognized phrase;

identifying a plurality of alternative phrases that may correspond to the spoken utterance, wherein the identifying of alternative phrases does not identify any alternative phrase that differs from the misrecognized phrase by only one words;

displaying the identified alternative phrases;

receiving from the speaker a selection of a displayed phrase; and

displaying the selected phrase in place of the misrecognized phrase.

6. A computer-readable medium having instructions for causing a continuous-speech recognition computer system to correct a spoken utterance that is misrecognized as a phrase having a plurality of words, by:

displaying the misrecognized phrase;

receiving from a speaker an indication to correct the misrecognized phrase;

identifying a plurality of alternative phrases that may correspond to the spoken utterance, wherein the identifying of alternative phrases does not identify any alternative phrase that can be corrected by selecting alternative words for words in the misrecognized phrase;

displaying the identified alternative phrases;

receiving from the speaker a selection of a displayed phrase; and

displaying the selected phrase in place of the misrecognized phrase.

8. A computer-readable medium having instructions for causing a continuous-speech recognition computer system to correct a spoken utterance that is misrecognized as a phrase having a plurality of words, by:

displaying the misrecognized phrase;

receiving from a speaker an indication to correct the misrecognized phrase;

identifying a plurality of alternative phrases that may correspond to the spoken utterance, wherein the identifying of alternative phrases does not include phrases with words whose correspondence between portions of the spoken utterance is approximately the same as the correspondence between words of the misrecognized phrase;

displaying the identified alternative phrases;

receiving from the speaker a selection of a displayed phrase; and

displaying the selected phrase in place of the misrecognized phrase.

9. A continuous-speech recognition computer system for correcting speech that is misrecognized, the speech being recognized as a plurality of words, comprising:

a display device;

a recognition component that receives speech and provides phrases with words representing the speech; and

a dictation editing component that displays words that represents a misrecognition of speech, that receives a selection of less than all of the displayed words representing a misrecognized phrase with two or more words, that after receiving the selection, displays on the display device a plurality of alternative phrases that may correspond to the misrecognized phrase, that receives a selection of a displayed phrase, and that displays the selected phrase in place of the misrecognized phrase on the display device.

10. A continuous-speech recognition computer system for correcting speech that is misrecognized, the speech being recognized as a plurality of words, comprising:

a display device;

a recognition component that receives speech and provides phrases with words representing the speech; and

a dictation editing component that displays words that represents a misrecognition of speech, that receives a selection of less than all of the displayed words representing a misrecognized phrase, that after receiving the selection, displays on the display device a plurality of alternative phrases that may correspond to the misrecognized phrase, that receives a selection of a displayed phrase, and that displays the selected phrase in place of the misrecognized phrase on the display device wherein the plurality of alternative phrases does not include any alternative phrase that can be corrected by selecting alternative words for words in the misrecognized phrase.