Adding speech recognition libraries to an existing program...

Details Adding speech recognition libraries to an existing program... Adding speech recognition libraries to an existing program...

1996-12-19

2003-11-25

Ingberg, T. D. (Department: 2124)

Data processing: software development, installation, and managem

Software program development tool

Linking

C704S231000, C704S275000

Reexamination Certificate

active

06654955

ABSTRACT:

FIELD OF THE INVENTION
The present invention relates to data processing systems, and more particularly, to enabling existing application programs for operation in speech recognition environments.
BACKGROUND OF THE INVENTION
Computer users have always yelled at their machines, but now computers are beginning to listen. The tap-tap-tap of the electronic workplace is being joined by a cacophony of conversation. Users can tell their computers to open and close files or perform other tasks by speaking a few words. Telephone callers can tell their carriers' computer systems to make a collect call or dial a business associate or suppliers.
Driving this move to listening computers is a one thousand per cent increase in microprocessor power, an accompanying price drop, and a new generation of voice-recognition devices. Another force bringing the power of voice to the desktop is the need to find an alternative input device for the keyboard and mouse. A logical replacement is a voice interface which allows a user to use a device available since birth. Speech recognition technology is available to the desktop user through the simple installation of a program and a microphone.
Typical prior art speech recognition operations occur in a single user, speech-dependent environment. This requires each speaker to train the speech recognizer with the user's voice patterns, during a process called “enrollment”. The system then maintains a profile for each speaker, who must identify themselves to the system in future recognition sessions. Typically speakers enroll via a local microphone in a low noise environment, speaking to the single machine on which the recognizer is resident. During the course of enrollment, the speaker is required to read a lengthy set of transcripts, so that the system can adjust itself to the peculiarities of each particular speaker. These systems require speakers to form each word in a halting and unnatural manner, pausing, between, each, and, every, word. This allows the speech recognizer to identify the voice pattern associated with each individual word using preceding, and following, silences to bound the words. The speech recognizer will typically have a single application for which it is trained.
More recently, a major advance occurred with the advent of speaker independent recognition systems that are capable of recognizing words from a continuous stream of conversational speech. This system requires no individualized speaker enrollment for effective use, unlike some speaker dependent systems which require speakers to be re-enrolled every four to six weeks, or require users carry a personalized plug-in cartridge to be understood by the system. With continuous speech recognition, no pauses between words are required, thus providing a more user friendly approach to the causal user of a speech recognition system. The growing familiarity and acceptance of speech has lead to more demand for speech aware applications. While applications have started to be designed for speech input, a large number of application programs have been written before this user friendly approach was available.
Consequently, it would be desirable to automatically enable existing application programs for operation in speech recognition environments without changing existing source code or recompiling the application programs.
SUMMARY OF THE INVENTION
This invention relates to a method and apparatus for enabling existing application programs for operation in speech recognition environments. Existing application programs written with a dynamically linked library or object library, with no speech recognition capability, are capable of accepting input from a speech recognition device without modification. This is accomplished by supplying an alternate dynamic library or object library that supports the same interface or objects used by the original program. The alternate library is written so that it is aware of and receives input from the speech recognition system. The alternate library then passes the input to the application program using the existing interfaces. The application program is unaware that the input comes from the speech recognition system instead of standard input devices such as keyboards or mouses.


REFERENCES:
patent: 4573187 (1986-02-01), Bui et al.
patent: 4677569 (1987-06-01), Nakano et al.
patent: 4766529 (1988-08-01), Nakano et al.
patent: 4776016 (1988-10-01), Hansen
patent: 4858170 (1989-08-01), DeWick, Sr. et al.
patent: 4910784 (1990-03-01), Doddington et al.
patent: 5157384 (1992-10-01), Greanias et al.
patent: 5179700 (1993-01-01), Aihara et al.
patent: 5252951 (1993-10-01), Tannenbaum et al.
patent: 5260697 (1993-11-01), Barrett et al.
patent: 5408582 (1995-04-01), Colier
patent: 5425128 (1995-06-01), Morrison
patent: 5442376 (1995-08-01), Tannenbaum et al.
patent: 5500920 (1996-03-01), Kupiec
patent: 5526407 (1996-06-01), Russell et al.
patent: 5530863 (1996-06-01), Hino
patent: 5555365 (1996-09-01), Selby et al.
patent: 5566248 (1996-10-01), Ulrich
patent: 5632002 (1997-05-01), Hashimoto et al.
patent: 5651094 (1997-07-01), Takagi et al.
patent: 5659665 (1997-08-01), Whelpley, Jr.
patent: 5664061 (1997-09-01), Andreshak et al.
Emiliani, Pier. IEEE Colloq. (1993) No. 103: Information Access for People. 11/1-11.3, 1993.*
McKiel, Frank. Audio enabled graphical user interface for the blind or visually impaired. Computer Applications to Assist Persons with Disabilities, 1992 Proceedings. 1992.*
Huang et al. “Microsoft Windows Highly intelligent speech recognizer: WHISPER.” Acoustics, Speech and Signal Processing. 1995.*
Schmandt et al. “Augmenting a window system with speech input.” Computer Magazine. 1990.*
Object-Orientewd DLLs Gen Kiyooka Beyond DOS, Dec. 1992.*
Object Oriented Analysis and Design with Applications, Grady Booch pp. 474-475, Sep. 1993.*
A Gentle Introduction to the VAX system, J.R. Hubbard pp. 117-118, 1987.*
Object-Oriented Software Construction, Bertrand Meyer pp. 339-444, Apr. 4, 1996.*
Microsoft Windows 95 Resource Kit, Microsoft Press pp. 1017-1037, Aug. 19, 1995.*
The Computer Speech Book Esther Schindler Chapters 2, 6,8,10,12,13, Feb. 1996.

Affiliated with

Also associated with

Rating

Adding speech recognition libraries to an existing program... does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Adding speech recognition libraries to an existing program..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Adding speech recognition libraries to an existing program... will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3132819