Speech recognition method and apparatus

Details Speech recognition method and apparatus Speech recognition method and apparatus

1998-03-31

2000-07-18

Hudspeth, David R.

Data processing: speech signal processing, linguistics, language

Speech signal processing

Recognition

704255, G10L 1508

Patent

active

060920424

ABSTRACT:
Speaker independent speech recognition is made highly accurately without setting any recognition unit, such as triphone, and by taking environment dependency of phonemes into considerations. A word dictionary unit 10 stores phoneme symbol series of a plurality of recognition subject words. A transition probability memory unit 20 stores transition probabilities associated with N.times.N mutual state transitions of N states in a given order to one another. An output probability memory unit 30 stores phoneme symbol output probabilities and feature vector output probabilities associated with the respective state transitions. A work comparing unit 40 calculates probabilities of sets of unknown input speech feature vector time series and hypothetical recognition subject words. A recognition result output unit 50 provides a highest probability word among all the recognition subject words as a result of recognition.

REFERENCES:
patent: 5230037 (1993-07-01), Giustiniani et al.
patent: 5598507 (1997-01-01), Kimber et al.
patent: 5608841 (1997-03-01), Tsuboka
patent: 5682501 (1997-10-01), Sharman
patent: 5721808 (1998-02-01), Minami et al.
patent: 5778341 (1998-07-01), Zeljkovic
patent: 6009390 (1997-03-01), Gupta et al.
Papers of the Institute of Electronics, Information and Communications Engineers, vol. J77-A, No. 2, Feb. 1994, "Applications in Speakers Without Instructors Using Whole Sound Ergodic HMM", pp. 112-119, (Published Feb. 25, 1994).
Technical Research Reports of the Institute of Electronics, Information and Communications Engineers [Audio] vol. 92, No. 274, SP92-75, "Applications in Speakers Without Instructors Using Whole Sound Ergordic HMM," pp. 15-20 (Published Oct. 21, 1992).
Technical Research Reports of the Institute of Electronics, Information and Communications Engineers [Audio] vol. 92, No. 410, SP92-129 Distinguishing Many Languages by means of Audio Using Ergodic HMM p. 49-66 (Issued Date: Jan. 19, 1993).
Papers of the Institute of Electronics, Information and Communications Engineers, vol. J77-A, No. 2, Feb. 1994, "Language Recognition by means of Audio Using Ergodic HMM and its State Sequences," p. 182-189, (Issue Date: Feb. 25, 1994).
L.R. Rabiner, B-H. Juang, Translated by Akira FURUI "Introduction to Speech Recognition" (Final Volume) Date of Publication: Nov. 1995, NTT Advanced Technology, pp. 135-138.
Rabiner et al., "Fundamentals of Speech Recognition", Prentice Hall, ISBN-0-13-055157-2, pp. 348, 458-460 1993.

Affiliated with

Also associated with

Rating

Speech recognition method and apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Speech recognition method and apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Speech recognition method and apparatus will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2047549