System and method for automating transcription services

• US 6,122,614 A
• Filed: 11/20/1998
• Issued: 09/19/2000
• Est. Priority Date: 11/20/1998
• Status: Expired due to Fees

• Current Assignee: Cambridge Software Systems Incorporated, Custom Speech USA, Inc.
• Original Assignee: Custom Speech USA, Inc.
• Inventors: Kahn, Jonathan, Flynn, Thomas P., Qin, Charles, Tippe, Robert J.
• Primary Examiner(s): Hudspeth, David R.
• Assistant Examiner(s): CHAWAN, VIJAY B
• Application Number:
  US09/197,313
• Time in Patent Office: 669 Days
• Field of Search: 704/260, 704/270, 704/275, 704/257, 704/277, 704/278, 704/235
• US Class Current: 704/235
• CPC Class Codes: G10L 15/063 Training
• CPC Class Codes: G10L 2015/0631 Creating reference templates; Clustering

Abstract

A system for substantially automating transcription services for multiple voice users including a manual transcription station, a speech recognition program and a routing program. The system establishes a profile for each of the voice users containing a training status which is selected from the group of enrollment, training, automated and stop automation. When the system receives a voice dictation file from a current voice user based on the training status the system routes the voice dictation file to a manual transcription station and the speech recognition program. A human transcriptionist creates transcribed files for each received voice dictation files. The speech recognition program automatically creates a written text for each received voice dictation file if the training status of the current user is training or automated. A verbatim file is manually established if the training status of the current user is enrollment or training and the speech recognition program is trained with an acoustic model for the current user using the verbatim file and the voice dictation file if the training status of the current user is enrollment or training. The transcribed file is returned to the current user if the training status of the current user is enrollment or training or the written text is returned if the training status of the current user is automated. An apparatus and method is also disclosed for simplifying the manual establishment of the verbatim file. A method for substantially automating transcription services is also disclosed.

Claims

1. A system for substantially automating transcription services for one or more voice users, said system comprising:

means for receiving a voice dictation file from a current user, said current user being one of said one or more voice users;

an audio player used to audibly reproduce said voice dictation file;

means for manually inputting and creating a transcribed file based on humanly perceived contents of said voice dictation file;

means for automatically converting said voice dictation file into written text;

means for manually editing a copy of said written text to create a verbatim text of said voice dictation file;

means for training said automatic speech converting means to achieve higher accuracy with said voice dictation file of current user; and

means for controlling the flow of said voice dictation file based upon a training status of said current user, whereby said controlling means sends said voice dictation file to at least one of said manual input means and said automatic speech converting means.

2. The invention according to claim 1 wherein said written text is at least temporarily synchronized to said voice dictation file, said manual editing means comprises:

means for sequentially comparing a copy of said written text with said transcribed file resulting in a sequential list of unmatched words culled from said copy of said written text, said sequential list having a beginning, an end and a current unmatched word, said current unmatched word being successively advanced from said beginning to said end;

means for incrementally searching for said current unmatched word contemporaneously within a first buffer associated with the speech recognition program containing said written text and a second buffer associated with said sequential list; and

means for correcting said current unmatched word in said second buffer, said correcting means including means for displaying said current unmatched word in a manner substantially visually isolated from other text in said copy of said written text and means for playing a portion of said synchronized voice dictation recording from said first buffer associated with said current unmatched word.

3. The invention according to claim 2 wherein said correcting means further includes means for alternatively viewing said current unmatched word in context within said copy of said written text.

4. The invention according to claim 3 wherein said sequential list and said written text each have a respective number of words, said accuracy rate determining means determines the ratio of said number of words in said sequential list to said number of words in said written text.

5. The invention according to claim 4 wherein said training status is based on said accuracy rate.

6. The invention according to claim 1 further including means for determining an accuracy rate for said current user.

7. The invention according to claim 1 wherein said automatic speech converting means comprises a preexisting speech recognition program intended for human interactive use, said automatic speech converting means includes means for automating responses to a series of interactive inquiries from said preexisting speech recognition program.

8. The invention according to claim 7 wherein said training means comprises a preexisting training portion of said preexisting speech recognition program intended for human interactive use, said training means includes means for automating responses to a series of interactive inquiries from said preexisting training portion of said preexisting speech recognition program.

9. The invention according to claim 1 wherein said training means comprises a preexisting training program intended for human interactive use, said training means includes means for automating responses to a series of interactive inquiries from said preexisting training program.

10. The invention according to claim 1 wherein said control means reads and modifies a user profile associated with said current user, each of said user profiles including said training status of said current user.

11. The invention according to claim 10 wherein said training status is selected from the group of pre-enrollment, enrollment, training, automation and stop automation.

12. The invention according to claim 11 when said training status is pre-enrollment said control means further includes means for creating a user identification and acoustic model within said automatic speech converting means.

13. The invention according to claim 11 when said training status is enrollment said control means routes said voice dictation file to said automatic speech converting means and said manual input means, routes said written text and said transcribed file to said manual editing means, routes said verbatim text to said training means and routes said transcribed file back to said current user as a finished text.

14. The invention according to claim 11 when said training status is training said control means routes said voice dictation file to said automatic speech converting means and said manual input means, routes said written text and said transcribed file to said manual editing means, routes said verbatim text to said training means and routes said transcribed file back to said current user as a finished text.

15. The invention according to claim 11 when said training status is automation said control means routes said voice dictation file only to said automatic speech converting means and routes said written text back to said current user as a finished text.

16. The invention according to claim 10 further including means for determining an accuracy rate for said current user, wherein said training status is based in part upon said accuracy rate.

17. The invention according to claim 1 further comprising means for manually selecting a specialized language model.

18. The invention according to claim 17 wherein said manual language model selection means is associated with said voice dictation file receiving means such that said current user can select said specialized language model.

19. The invention according to claim 1 wherein said voice dictation file receiving means is implemented by a general purpose computer.

20. The invention according to claim 19 wherein said general purpose computer implementing said voice dictation file receiving means is remotely located, said voice dictation file receiving means is operably connected to said audio player and said automatic speech conversion means by a computer network.

21. The invention according to claim 19 wherein said general purpose computer implementing said voice dictation file receiving means is remotely located, said general purpose computer further including means for transferring said voice dictation file via the Internet.

22. The invention according to claim 1 wherein said audio player is implemented by a general purpose computer.

23. The invention according to claim 22 wherein said general purpose computer implementing said audio player is remotely located, said audio player is operably connected to said voice dictation file receiving means and said automatic speech conversion means by a computer network.

24. The invention according to claim 22 wherein said general purpose computer implementing said audio player is remotely located, said general purpose computer further including means for receiving said voice dictation file and transmitting said transcribed text via the Internet.

25. An apparatus for substantially simplifying the production of a verbatim text of a voice dictation recording based upon a transcribed file produced by a human transcriptionist and a written text produced by a speech recognition program, wherein said written text is at least temporarily synchronized to said voice dictation recording, said apparatus comprising:

means for sequentially comparing a copy of said written text with said transcribed file resulting in a sequential list of unmatched words culled from said copy of said written text, said sequential list having a beginning, an end and a current unmatched word, said current unmatched word being successively advanced from said beginning to said end;

means for incrementally searching for said current unmatched word contemporaneously within a first buffer associated with the speech recognition program containing said written text and a second buffer associated with said sequential list; and

means for correcting said current unmatched word in said second buffer, said correcting means including means for displaying said current unmatched word in a manner substantially visually isolated from other text in said copy of said written text and means for playing a portion of said synchronized voice dictation recording from said first buffer associated with said current unmatched word.

26. The invention according to claim 25 wherein said correcting means further includes means for alternatively viewing said current unmatched word in context within said copy of said written text.

27. The invention according to claim 25 wherein said manner substantially visually isolated from other text can be manually selected from the group containing word-by-word display, sentence-by-sentence display, and said current unmatched word display.

28. The invention according to claim 25 further including means for determining an accuracy rate for said current user.

29. The invention according to claim 28 wherein said sequential list and written text each have a respective number of words, said accuracy rate determining means determines the ratio of said number of words in said sequential list to said number of words in said written text.

30. A method for automating transcription services for one or more voice users in a system including a manual transcription station and a speech recognition program, said method comprising the steps of:

establishing a profile for each of the voice users, the profile containing a training status;

receiving a voice dictation file from a current voice user;

choosing the training status of the current voice user from the group of enrollment, training, automated and stop automation;

routing the voice dictation file to at least one of the manual transcription station and the speech recognition program based on the training status;

receiving the voice dictation file in at least one of the manual transcription station and the speech recognition program;

creating a transcribed file at the manual transcription station for each received voice dictation file;

automatically creating a written text with the speech recognition program for each received voice dictation file if the training status of the current user is training or automated;

manually establishing a verbatim file if the training status of the current user is enrollment or training;

training the speech recognition program with an acoustic model for the current user using the verbatim file and the voice dictation file if the training status of the current user is enrollment or training;

returning the transcribed file to the current user if the training status of the current user is enrollment or training; and

returning the written text to the current user if the training status of the current user is automated.

31. The invention according to claim 30 wherein said step of manually establishing a verbatim file includes the sub-steps of:

assisting an operator to establish the verbatim file if the training status of the current user is training by;

sequentially comparing a copy of the written text with the transcribed file resulting in a sequential list of unmatched words culled from the copy of the written text, the sequential list having a beginning, an end and a current unmatched word, the current unmatched word being successively advanced from the beginning to the end;

incrementally searching for the current unmatched word contemporaneously within a first buffer associated with the speech recognition program containing the written text and a second buffer associated with the sequential list; and

displaying the current unmatched word in a manner substantially visually isolated from other text in the copy of the written text and playing a portion of the synchronized voice dictation recording from the first buffer associated with the current unmatched word; and

correcting the current unmatched word to be a verbatim representation of the portion of the synchronized voice dictation recording.

32. A method for assisting an operator to establish a verbatim file from a written text created by a speech recognition program and a transcribed file created by a human transcriptionist, the method comprising the steps of:

sequentially comparing a copy of the written text with the transcribed file resulting in a sequential list of unmatched words culled from the copy of the written text, the sequential list having a beginning, an end and a current unmatched word, the current unmatched word being successively advanced from the beginning to the end;

incrementally searching for the current unmatched word contemporaneously within a first buffer associated with the speech recognition program containing the written text and a second buffer associated with the sequential list; and

displaying the current unmatched word in a manner substantially visually isolated from other text in the copy of the written text and playing a portion of the synchronized voice dictation recording from the first buffer associated with the current unmatched word; and

correcting the current unmatched word to be a verbatim representation of the portion of the synchronized voice dictation recording.