Word recognition using hidden control neural architecture

Author

Levin, Esther

Author_Institution

AT&T Bell Lab., Murray Hill, NJ, USA

fYear

1990

fDate

3-6 Apr 1990

Firstpage

433

Abstract

Neural networks are used to model nonlinear and time-varying systems. The proposed model attempts to cope with the time variability systems by adding an undetermined control input which modulates the mapping implemented by the network. The network architecture proposed, the hidden control neural network (HCNN), combines nonlinear prediction of conventional neural networks with hidden Markov modeling. This network is trained using an algorithm that is based on back-propagation and segmentation algorithms for estimating the unknown control together with the network´s parameters. The HCNN approach is evaluated on multispeaker recognition of connected digits, yielding a word accuracy of 99.3%

Keywords

Markov processes; neural nets; speech recognition; time-varying systems; back-propagation; connected digits; hidden Markov modeling; hidden control neural architecture; hidden control neural network; multispeaker recognition; nonlinear prediction; segmentation algorithms; speech recognition; time-varying systems; training algorithm; undetermined control input; word accuracy; word recognition; Hidden Markov models; Linear systems; Modeling; Multi-layer neural network; Neural networks; Nonlinear control systems; Predictive models; Signal mapping; Speech; Time varying systems;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1990. ICASSP-90., 1990 International Conference on

Conference_Location

Albuquerque, NM

ISSN

1520-6149

Type

conf

DOI

10.1109/ICASSP.1990.115740

Filename

115740