Neural Networks Applied to Speech Recognition

Hiroaki SAKOE

Neural Networks Applied to Speech Recognition

Hiroaki SAKOE

Full Text Views

0

Cite this

Summary :

Applications of neural networks are prevailing in speech recognition research. In this paper, first, suitable role of neural network (mainly back-propagation based multi-layer type) in speech recognition, is discussed. Considering that speech is a long, variable length, structured pattern, a direction, in which neural network is used in cooperation with existing structural analysis frameworks, is recommended. Activities are surveyed, including those intended to cooperatively merge neural networks into dynamic programming based structural analysis framework. It is observed that considerable efforts have been paid to suppress the high nonlinearity of network output. As far as surveyed, no experiment in real field has been reported.

Publication: IEICE TRANSACTIONS on Fundamentals Vol.E75-A No.5 pp.546-551

Publication Date: 1992/05/25

Publicized

Online ISSN

DOI

Type of Manuscript: Special Section INVITED PAPER (Special Section on Nonlinear Dynamics--Adaptive, Learning and Neural Systems--)

Category

Cite this

Copy

Hiroaki SAKOE, "Neural Networks Applied to Speech Recognition" in IEICE TRANSACTIONS on Fundamentals, vol. E75-A, no. 5, pp. 546-551, May 1992, doi: .
Abstract: Applications of neural networks are prevailing in speech recognition research. In this paper, first, suitable role of neural network (mainly back-propagation based multi-layer type) in speech recognition, is discussed. Considering that speech is a long, variable length, structured pattern, a direction, in which neural network is used in cooperation with existing structural analysis frameworks, is recommended. Activities are surveyed, including those intended to cooperatively merge neural networks into dynamic programming based structural analysis framework. It is observed that considerable efforts have been paid to suppress the high nonlinearity of network output. As far as surveyed, no experiment in real field has been reported.
URL: https://global.ieice.org/en_transactions/fundamentals/10.1587/e75-a_5_546/_p

Copy

@ARTICLE{e75-a_5_546,
author={Hiroaki SAKOE, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={Neural Networks Applied to Speech Recognition},
year={1992},
volume={E75-A},
number={5},
pages={546-551},
abstract={Applications of neural networks are prevailing in speech recognition research. In this paper, first, suitable role of neural network (mainly back-propagation based multi-layer type) in speech recognition, is discussed. Considering that speech is a long, variable length, structured pattern, a direction, in which neural network is used in cooperation with existing structural analysis frameworks, is recommended. Activities are surveyed, including those intended to cooperatively merge neural networks into dynamic programming based structural analysis framework. It is observed that considerable efforts have been paid to suppress the high nonlinearity of network output. As far as surveyed, no experiment in real field has been reported.},
keywords={},
doi={},
ISSN={},
month={May},}

Copy

TY - JOUR
TI - Neural Networks Applied to Speech Recognition
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 546
EP - 551
AU - Hiroaki SAKOE
PY - 1992
DO -
JO - IEICE TRANSACTIONS on Fundamentals
SN -
VL - E75-A
IS - 5
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - May 1992
AB - Applications of neural networks are prevailing in speech recognition research. In this paper, first, suitable role of neural network (mainly back-propagation based multi-layer type) in speech recognition, is discussed. Considering that speech is a long, variable length, structured pattern, a direction, in which neural network is used in cooperation with existing structural analysis frameworks, is recommended. Activities are surveyed, including those intended to cooperatively merge neural networks into dynamic programming based structural analysis framework. It is observed that considerable efforts have been paid to suppress the high nonlinearity of network output. As far as surveyed, no experiment in real field has been reported.
ER -