Using parallel MLPs as labelers for multiple codebook HMMs

Authors:
Philippe Le Cerf;Dirk Van Compernolle
Affiliations:
K. U. Leuven, E.S.A.T., Heverlee, Belgium;K. U. Leuven, E.S.A.T., Heverlee, Belgium
Venue:
ICASSP'93 Proceedings of the 1993 IEEE international conference on Acoustics, speech, and signal processing: plenary, special, audio, underwater acoustics, VLSI, neural networks - Volume I
Year:
1993

Citing 3
Cited 0

Links Between Markov Models and Multilayer Perceptrons

IEEE Transactions on Pattern Analysis and Machine Intelligence
Connectionist probability estimation in the DECIPHER speech recognition system

ICASSP'92 Proceedings of the 1992 IEEE international conference on Acoustics, speech and signal processing - Volume 1
A speech recognizer using radial basis function neural networks in an HMM framework

ICASSP'92 Proceedings of the 1992 IEEE international conference on Acoustics, speech and signal processing - Volume 1

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, we investigate the use of MLPs as labelers for a discrete parameter HMM system. We introduce a number of strategies of which the Multi-MLP approach, which uses parallel MLPs for separate parameter sets, is the most promising. The performance of the new system is just as good as that of a classical discrete parameter HMM system (using multiple Euclidean VQs), but needs fewer HMM parameters (80 compared with 330 per state). Therefore, Multi-MLP labeling is much more efficient than Euclidean labeling.