Data processing: speech signal processing – linguistics – language – Speech signal processing – Recognition
Reexamination Certificate
1999-06-30
2002-04-09
Dorvil, Richemond (Department: 2741)
Data processing: speech signal processing, linguistics, language
Speech signal processing
Recognition
C714S738000, C714S738000
Reexamination Certificate
active
06370503
ABSTRACT:
BACKGROUND OF THE INVENTION
The field of the invention is methods and apparatuses for continuous speech recognition. More particularly, the invention relates to methods for presenting the user with control over speech recognition accuracy.
Speech recognition systems currently exist which enable a user to continuously dictate speech into a microphone, whereupon that speech is transformed to text by a computer, and the text is displayed on the computer's monitor. In a continuous speech recognition system designed to decode speech without pauses between words, it is reasonable to expect that errors of decoding will take place and will have to be corrected by the user. Often, a particular type of error will continuously recur during a particular dictation session. For example, the system may repeatedly mis-recognize a particular word, such as “and,” for example.
Prior-art speech recognition systems receive and transcribe received speech based on various system parameters and databases, such as audio levels, topic libraries, and speaker-dependent training databases, for example. In some systems, a user trained on the system may be able to determine what type of recognition accuracy problem is occurring, and adjust or change the various parameters or databases in order to improve the recognition accuracy of the user's speech. Many users, however, may not have the inclination or patience to learn how to identify recognition accuracy problems or how to adjust the system parameters to improve upon those problems. In addition, some users may prefer not to monitor the text being generated during a speech recognition session, and may not realize that recognition problems are occurring. Both of these types of users may become frustrated and decide not to use the system.
What is needed is a method and apparatus for transcribing dictated speech that enables a user easily to change system parameters or user behaviors in order to improve the recognition accuracy of the system. What is further needed is a method and apparatus for transcribing dictated speech that can identify when speech recognition problems are occurring, and can assist the user in identifying the particular type of problem.
SUMMARY OF THE INVENTION
The present invention includes an apparatus and method for transcribing speech that receives input speech, and transcribes the input speech into transcribed speech that includes a sequential listing of words in text format. During transcription, the method and apparatus monitors the accuracy of the transcribed speech, and determines whether the accuracy of the transcribed speech is sufficient. If the accuracy is not sufficient, the method and apparatus automatically activates a speech recognition improvement tool. Using the speech recognition improvement tool, the method and apparatus improves the accuracy of the transcribing process.
The method is carried out by a transcription apparatus which includes at least a computer, a monitor, and a microphone. The method can be executed by a machine that executes a plurality of code sections of a computer program that is stored on a machine readable storage.
A general object of the invention is to provide a transcription system user with the ability and the information to easily change system parameters and/or user behaviors in order to improve the recognition accuracy of the system. Another object of the invention is to provide the user with this ability automatically.
REFERENCES:
patent: 5684924 (1997-11-01), Stanley et al.
patent: 5715469 (1998-02-01), Arning
patent: 5829000 (1998-10-01), Huang et al.
patent: 5855000 (1998-12-01), Waibel et al.
patent: 5864805 (1999-01-01), Chen et al.
patent: 5909667 (1999-06-01), Leontiades et al.
patent: 5950160 (1999-09-01), Rozak
patent: 6064959 (2000-05-01), Young et al.
Egger Hans
Keller Arthur
Lewis James R.
Ortega Kerry A.
Vanbuskirk Ronald E.
Abebe Daniel
Akerman & Senterfitt
Dorvil Richemond
International Business Machines Corp.
LandOfFree
Method and apparatus for improving speech recognition accuracy does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for improving speech recognition accuracy, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for improving speech recognition accuracy will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2867134