Neural learning for articulatory speech synthesis under different statistical characteristics of acoustic input patterns
Küçük Resim Yok
Tarih
2003
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
PERGAMON-ELSEVIER SCIENCE LTD
Erişim Hakkı
info:eu-repo/semantics/closedAccess
Özet
Input data representation is highly decisive in neural learning in terms of convergence. In this paper, within an analytical and statistical framework, the effect of the distribution characteristics of the input pattern vectors on the performance of the back-propagation (BP) algorithm is established for a function approximation problem, where parameters of an articulatory speech synthesizer are estimated from acoustic input data. The aim is to determine the optimum statistical characteristics of the acoustic input patterns in order to improve neural learning. Improvement is obtained through a modification of the statistical characteristics of the input data, which reduces effectively the occurrence of node saturation in the hidden layer. (C) 2002 Elsevier Science Ltd. All rights reserved.
Açıklama
Anahtar Kelimeler
back-propagation algorithm, statistical dependency of neural learning, articulatory speech synthesizer
Kaynak
COMPUTERS & ELECTRICAL ENGINEERING
WoS Q Değeri
Q4
Scopus Q Değeri
Q1
Cilt
29
Sayı
6
