Pattern recognition using a predictive neural network

Details Pattern recognition using a predictive neural network Pattern recognition using a predictive neural network

1993-11-30

1997-12-09

Knepper, David D.

395 25, 395 247, G10L 506

Patent

active

056968779

ABSTRACT:
Input feature vectors (a(t)) is considered a pattern selected from a plurality of reference patterns which represent categories of recognition objects. Each reference pattern is defined by a sequence of state models, successively supplied with the time sequence of the input feature vectors and with a sequence of preceding state vectors (h(t, s, n)). The sequence of the state models produces a time sequence of predicted feature vectors (A(t+1, s, n) and a sequence of new state vectors (h(t+1, s, n)). The recognized pattern is selected from one of the reference patterns that minimizes a prediction error between the time sequence of the input feature vectors and the time sequence of the predicted feature vectors. The prediction error is calculated by using a dynamic programming algorithm. Training of the reference pattern is carried out by a gradient descent method such as back-propagation technique.

REFERENCES:
Lippmann, "An Introduction to Computing with Neural Nets", IEEE ASSP Mazagine, Apr. 1987, pp. 4-22.
Waibel et al., "Phoneme Recognition Using Time-Delay Neural Networks", IEEE Trans. on ASSp, vol. 37, No. 3, Mar. 1989, pp. 328-339.
Cambridge University Engineering Department Technical Report CUED/F-INFENG/TR27 1988, Cambridge GB, pp. 1-13; F. Fallside: `On the analysis of linear predictive data such as speech by a class of single layer connectionist models`.
International Conference On Acoustics Speech And Signal Processing, vol. 1, 23 May 1989, Glascow Scotland U.K. pp. 29-32; H. Sakae et al; `Speaker-independent word recognition using dynamic programming neutral networks`.
International Conference On Acoustics Speech And Signal Processing, vol. 1, 3 Apr. 1990, Albuquerque New Mexico USA, pp. 441-444; ISO, K.-I.; Watanabe, T.P: `Speaker-independent word recognition using a neural predition model.`
International Conference On Acoustics Speech And Signal Processing, vol. 1, 3 Apr. 1990, Albuguerque New Mexico USA, pp. 437-440; Tebelskis, J: Waibel, A: `Large vocabulary recognition using linked predictive neural networks.`
International Conference On Acoustics Speech And Signal Processing, vol. 1, 11 Apr. 1988, New York, USA, pp. 107-110; Waibel, A. et al: `Phoneme recognition: neural networks vs. Hidden Markov Model`.
L.R. Rabiner and B.H. Juang, "An Introduction to Hidden Markov Models," IEEE ASSP Mazagine, Jan. 1986, pp. 4-16.

Affiliated with

Also associated with

Rating

Pattern recognition using a predictive neural network does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Pattern recognition using a predictive neural network, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Pattern recognition using a predictive neural network will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-1614745