Automatic management system for speech recognition processes

• US 5,425,128 A
• Filed: 05/29/1992
• Issued: 06/13/1995
• Est. Priority Date: 05/29/1992
• Status: Expired due to Fees

• Current Assignee: Sunquest Information Systems, Inc. (Roper Technologies, Inc.)
• Original Assignee: Sunquest Information Systems, Inc. (Roper Technologies, Inc.)
• Inventors: Morrison, Robert L.
• Primary Examiner(s): MacDonald, Allen R.
• Assistant Examiner(s): Doerrler, Michelle
• Application Number:
  US07/890,915
• Time in Patent Office: 1,110 Days
• Field of Search: 395/2.52, 395/2.53, 395/2.66, 395/2.55, 395/2.79, 395/2.84, 381/43
• US Class Current: 704/243
• CPC Class Codes: G10L 15/06 Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice G10L15/14 takes precedence
• CPC Class Codes: G10L 15/22 Procedures used during a speech recognition process, e.g. man-machine dialogue
• CPC Class Codes: G10L 2015/228 of application context

Abstract

In a computer system including an input-output terminal and an application program having a vocabulary file of acceptable input words and keystroke characters associated with the input words, the vocabulary file and corresponding keystroke characters are automatically retrieved by an operator request through the action of a host application program. Prior to executing a speech recognition process, the retrieved vocabulary file and corresponding keystroke characters are automatically assembled by the same host application program into a syntax file according to syntax rules of a speech recognition program. The resulting vocabulary/syntax file then is compiled into a format useable by the speech recognition process program. The compiled vocabulary/syntax file is used to automatically prompt the operator to speak the various vocabulary words, causing an analog-to-digital conversion circuit to produce digital template codes that are assembled into a speech template file. The compiled vocabulary/syntax file and the speech template file are sent to and stored with the application program, and are automatically downloaded to the speech recognition program to effectuate speech-input execution of the host application program.

Claims

1. An automatic speech recognition management system in a computer system including a speech recognition station and first storing means for storing an application program having an uncompiled vocabulary file of acceptable input words and corresponding characters associated with the input words, the automatic speech recognition management system comprising in combination:

(a) means for entering an operator identifier corresponding to a particular operator;

(b) means for executing the application program in response to the operator identifier to retrieve the uncompiled vocabulary file and corresponding characters;

(c) means for assembling the retrieved uncompiled vocabulary file and corresponding characters into a syntax file according to syntax rules of a speech recognition program associated with the speech recognition station;

(d) means associated with the speech recognition station for compiling the syntax file by means of a compiler associated with the speech recognition program to produce a compiled vocabulary file corresponding to the operator identifier and including the input words and corresponding character codes selected from a character code list;

(e) means for operating on the compiled vocabulary file to prompt an operator to speak the input words;

(f) analog-to-digital conversion circuit means for producing a digital signal representing the spoken input words;

(g) means responsive to the digital signal for producing therefrom a speech template file containing digital template codes corresponding to the input words;

(h) second storing means associated with the application program for storing the compiled vocabulary file and the speech template file;

(i) means for transmitting the compiled vocabulary file and the speech template file to the second storing means; and

(j) means for downloading the compiled vocabulary file and the speech template file from the second storing means to the speech recognition program in response to the operator identifier to enable the operator to speak words of the compiled vocabulary file to effectuate execution of the application program.

2. An automatic speech recognition management system in a computer system including a speech recognition station and first storing means for storing a plurality of application programs each having an uncompiled vocabulary file of acceptable input words and corresponding characters associated with the input words, the automatic speech recognition management system comprising in combination:

(a) means for entering an operator identifier corresponding to a particular operator and an application program identifier into the computer system;

(b) means for retrieving the uncompiled vocabulary file and corresponding characters of an application program selected by the entered application program identifier;

(c) means for assembling the retrieved uncompiled vocabulary file and corresponding characters into a syntax file according to syntax rules of a speech recognition program associated with the speech recognition station;

(d) means associated with the speech recognition station for compiling the syntax file by means of a compiler associated with the speech recognition program to produce a compiled vocabulary file including the input words and a plurality of character codes selected from a character code list, the character codes corresponding, respectively, to the characters;

(e) means for operating on the compiled vocabulary file to prompt the operator to speak the input words;

(f) an analog-to-digital conversion circuit producing a digital signal representing the spoken input words;

(g) means responsive to the digital signal for producing therefrom a speech template file associated with the operator identifier and containing digital template codes corresponding to the input words, respectively;

(h) means for transmitting the compiled vocabulary file and the speech template file to the application program;

(i) second storing means for storing the compiled vocabulary file and the speech template file received from the transmitting means; and

(j) means for downloading the compiled vocabulary file and the speech template file to the speech recognition program in response to the operator identifier and the application program identifier, to enable the operator to speak words of the compiled vocabulary file to effectuate execution of the application program.

3. The automatic speech recognition management system of claim 2 wherein the computer system includes a host computer, and the host computer includes the first storing means.

4. The automatic speech recognition management system of claim 3 wherein the speech recognition station includes a personal computer coupled to the host computer, the entering means including a keyboard included in the personal computer.

5. The automatic speech recognition management system of claim 4 wherein the speech recognition program is stored in the personal computer.

6. The automatic speech recognition management system of claim 4 wherein the characters are keystroke characters associated with the keyboard.

7. The automatic speech recognition management system of claim 2 including means for testing a vocabulary update flag associated with the application program to determine if the vocabulary file and corresponding characters have been updated, the transmitting means transmitting the compiled vocabulary file and speech template file to the speech recognition program if the vocabulary file and corresponding characters have not been updated.

8. The automatic speech recognition management system of claim 7 including means for testing an operator enrollment flag associated with the application program to determine if the operator has been enrolled for the application program since a most recent update of the vocabulary file thereof.

9. The automatic speech recognition management system of claim 7 wherein the assembling means includes means for setting up a software structure for entry of a plurality of uncompiled input words and corresponding keystroke characters and means for sequentially entering the uncompiled input words and associated characters into the software structure according to syntax rules of the speech recognition program.

10. An automatic speech recognition management system in a computer system including a speech recognition station and means for storing an application program having an uncompiled vocabulary file of acceptable input words and corresponding characters associated with the input words, the automatic speech recognition management system comprising in combination:

(a) means for entering an operator identifier corresponding to a particular operator;

(b) means for executing the application program to retrieve the uncompiled vocabulary file and corresponding characters;

(c) means for assembling the retrieved uncompiled vocabulary file and corresponding characters into a syntax file according to syntax rules of a speech recognition program associated with the speech recognition station;

(d) means associated with the speech recognition station for compiling the syntax file by means of a compiler associated with the speech recognition program to produce a compiled vocabulary file corresponding to the operator identifier and including the input words and corresponding character codes selected from a character code list;

(e) means for operating on the compiled vocabulary file to prompt an operator to speak the input words;

(f) analog-to-digital conversion circuit means for producing digital signals representing input words spoken by the operator;

(g) means responsive to the digital signals for producing therefrom a speech template file containing digital template codes corresponding to the input words;

(h) means for operating on a digital signal produced in response to speaking of a word by the operator and on data of the speech template file and the compiled vocabulary file to select the character code that represents the word if the word is included in the compiled vocabulary file; and

(i) means for transmitting the selected character code to the application program in response to the operator identifier to effectuate execution thereof.

11. In a computer system including a speech recognition station and means for entering operator identifiers and application program identifiers into the computer system, and means for storing a plurality of application programs each having an uncompiled vocabulary file of acceptable input words and corresponding characters associated with the input words, an automatic method of speech recognition management comprising the steps of:

(a) entering an operator identifier and an application program identifier into the computer system via the entering means;

(b) determining if a compiled vocabulary corresponding to the application program identifier exists, and if it does not exist, retrieving an uncompiled vocabulary file including input words and corresponding keystroke characters of an application program corresponding to the application program identifier and performing steps (c) and (d) before step (e);

(c) assembling the retrieved uncompiled vocabulary file and corresponding keystroke characters into a syntax file according to syntax rules of a speech recognition program stored in the computer system;

(d) compiling the syntax file by means of a compiler associated with the speech recognition program to produce a compiled vocabulary file corresponding to the entered application program identifier and including the input words of the application program and corresponding character codes selected from a character code list;

(e) using the compiled vocabulary file to prompt the operator to speak the various vocabulary words;

(f) producing a digital signal representing the spoken words by means of an analog-to-digital conversion circuit and producing therefrom a speech template file corresponding to the operator identifier and containing digital template codes corresponding to the input words of the compiled vocabulary file;

(g) transmitting the compiled vocabulary file and the speech template file to the application program and storing the compiled vocabulary file and the speech template file; and

(h) downloading the compiled vocabulary file and the speech template file to the speech recognition program in response to the application program identifier and the operator identifier to enable the operator to speech-input spoken words of the compiled vocabulary file to effectuate execution of the application program.

12. The method of claim 11 including testing a vocabulary update flag associated with the application program to determine if the vocabulary file and corresponding characters have been updated before performing step (b).

13. The method of claim 12 including testing an operator enrollment flag associated with the application program to determine if an operator to which the operator identifier corresponds has been enrolled for the application program since a most recent update of the vocabulary file thereof before performing step (e).

14. An automatic speech recognition management system in a computer system including a plurality of speech recognition stations and means for storing a plurality of application programs each having a vocabulary file of acceptable input words, the automatic speech recognition management system comprising in combination:

(a) means for entering an application program identifier of a desired application program and an operator identifier of a particular person requesting access to the desired application program into one of the speech recognition stations;

(b) means responsive to the application program identifier for retrieving a first file of information uniquely associated with the desired application program;

(c) means responsive to the operator identifier for retrieving a second file of information uniquely associated with the requesting person;

(d) means for operating the computer system to use information from the first file to determine a plurality of words that need to be spoken by the particular person into the speech recognition station to effectuate desired execution of the desired application program; and

(e) means for operating the computer system to use information from the second file to automatically augment recognition of the plurality of words spoken into the speech recognition station by the particular person.

15. An automatic speech recognition management system in a computer system including a speech recognition station, a first portion of the computer system storing a plurality of application programs each having an uncompiled vocabulary file of acceptable input words and corresponding characters associated with the input words, the automatic speech recognition management system comprising in combination:

(a) an input device for entering an operator identifier corresponding to a particular operator and an application program identifier into the computer system;

(b) a second portion of the computer system retrieving the uncompiled vocabulary file and corresponding characters of an application program selected by the entered application program identifier;

(c) a third portion of the computer system assembling the retrieved uncompiled vocabulary file and corresponding characters into a syntax file according to syntax rules of a speech recognition program associated with the speech recognition station;

(d) a first portion of the speech recognition station configured to compile the syntax file by means of a compiler associated with the speech recognition program to produce a compiled vocabulary file including the input words and a plurality of character codes selected from a character code list, the character codes corresponding, respectively, to the characters;

(e) a second portion of the speech recognition station operating on the compiled vocabulary file to prompt the operator to speak various input words;

(f) an analog-to-digital conversion circuit producing a digital signal representing the spoken input words;

(g) a third portion of the speech recognition station responsive to the digital signal producing therefrom a speech template file associated with the operator identifier and containing digital template codes corresponding to the input words, respectively;

(h) a fourth portion of the speech recognition station transmitting the compiled vocabulary file and the speech template file to the application program;

(i) a fourth portion of the computer system storing the compiled vocabulary file and the speech template file received from the fourth portion of the speech recognition station; and

(j) a fifth portion of the computer system downloading the compiled vocabulary file and the speech template file to the speech recognition program in response to the operator identifier and the application program identifier, to enable the operator to speak words of the compiled vocabulary file to effectuate execution of the application program.

16. In a computer system including a speech recognition station and an input device for entering operator identifiers and application program identifiers into the computer system, the computer system storing a plurality of application programs each having an uncompiled vocabulary file of acceptable input words and corresponding characters associated with the input words, an automatic method of speech recognition management comprising the steps of:

(a) entering an operator identifier and an application program identifier into the computer system;

(b) determining if a compiled vocabulary corresponding to the application program identifier exists, and if it does not exist, retrieving an uncompiled vocabulary file including input words and corresponding keystroke characters of an application program corresponding to the application program identifier and performing steps (c) and (d) before step (e);

(c) assembling the retrieved uncompiled vocabulary file and corresponding keystroke characters into a syntax file according to syntax rules of a speech recognition program stored in the computer system;

(d) compiling the syntax file by means of a compiler associated with the speech recognition program to produce a compiled vocabulary file corresponding to the entered application program identifier and including the input words of the application program and corresponding character codes selected from a character code list;

(e) using the compiled vocabulary file to prompt the operator to speak the various vocabulary words;

(f) producing a digital signal representing the spoken words by means of an analog-to-digital conversion circuit and producing therefrom a speech template file corresponding to the operator identifier and containing digital template codes corresponding to various input words of the compiled vocabulary file;

(g) transmitting the compiled vocabulary file and the speech template file to the application program and storing the compiled vocabulary file and the speech template file; and

(h) downloading the compiled vocabulary file and the speech template file to the speech recognition program in response to the application program identifier and the operator identifier to enable the operator to speech-input spoken words of the compiled vocabulary file to effectuate execution of the application program.

17. An automatic speech recognition management system in a computer system including a speech recognition station, the computer system storing a plurality of application programs each having a vocabulary file of acceptable input words, the automatic speech recognition management system comprising in combination:

(a) an input device for entering an application program identifier of a desired application program and an operator identifier of a particular person requesting access to the desired application program into the speech recognition station;

(b) a first portion of the computer system retrieving a first file of information uniquely associated with the desired application program in response to the application program identifier;

(c) a second portion of the computer system retrieving a second file of information uniquely associated with the requesting person in response to the operator identifier;

(d) a third portion of the computer system using information from the first file to assemble, according to rules of the speech recognition station, a plurality of words that need to be spoken by the particular person into the speech recognition station to effectuate desired execution of the desired application program; and

(e) a fourth portion of the computer system using information from the second file to automatically augment recognition of the plurality of words spoken into the speech recognition station by the particular person.

18. An automatic speech recognition management system in a computer system including a speech recognition station, a first portion of the computer system storing a plurality of application programs each having a vocabulary file of acceptable input words and corresponding characters associated with the input words, the automatic speech recognition management system comprising in combination:

(a) an input device for entering an operator identifier corresponding to a particular operator and an application program identifier into the computer system;

(b) a second portion of the computer system retrieving the vocabulary file and corresponding characters of an application program selected by the entered application program identifier;

(c) a third portion of the computer system assembling the retrieved vocabulary file and corresponding characters into a software language modeling structure according to language modeling rules of a speech recognition program associated with the speech recognition station;

(d) a first portion of the speech recognition station operating on the software language modeling structure to prompt the operator to speak various input words;

(e) an analog-to-digital conversion circuit producing a digital signal representing the spoken input words;

(f) a second portion of the speech recognition station responsive to the digital signal producing therefrom phonetic information associated with the operator identifier;

(g) a third portion of the speech recognition station transmitting the software language modeling structure and the phonetic information to the application program;

(h) a fourth portion of the computer system storing the software language modeling structure and the phonetic information received from the third portion of the speech recogniton station; and

(i) a fifth portion of the computer system downloading the software language modeling structure and the phonetic information to the speech recognition program in response to the operator identifier and the application program identifier, to enable the operator to speak words of the vocabulary file to effectuate execution of the application program.

19. An automatic speech recognition management system in a computer system including a speech recognition station, a first portion of the computer system storing a plurality of application programs each having a vocabulary file of acceptable input words and corresponding characters associated with the input words, the automatic speech recognition management system comprising in combination:

(a) an input device for entering an application program identifier into the computer system;

(b) a second portion of the computer system retrieving the vocabulary file and corresponding characters of an application program selected by the entered application program identifier;

(c) a third portion of the computer system assembling the retrieved vocabulary file and corresponding characters into a software language modeling structure according to language modeling rules of a speech recognition program associated with the speech recognition station;

(d) a first portion of the speech recognition station operating on the software language modeling structure to prompt an operator to speak various input words;

(e) an analog-to-digital conversion circuit producing a digital signal representing the spoken input words;

(f) a second portion of the speech recognition station responsive to the digital signal producing therefrom phonetic information;

(g) a third portion of the speech recognition station transmitting the software language modeling structure and the phonetic information to the application program;

(h) a fourth portion of the computer system storing the software language modeling structure and the phonetic information received from the third portion of the speech recognition station; and

(i) a fifth portion of the computer system downloading the software language modeling structure and the phonetic information to the speech recognition program in response to the application program identifier, to enable the operator to speak words of the vocabulary file to effectuate execution of the application program.