Application of Feature Extraction in Text-to-Speech Processing

0404196 - UIVT-O 20010057 RIV AT eng C - Conference Paper (international conference)
Šebesta, Václav - Tučková, Jana
Application of Feature Extraction in Text-to-Speech Processing.
Artificial Neural Nets and Genetic Algorithms. Proceedings of the International Conference. Wien: Springer, 2001 - (Kůrková, V.; Steele, N.; Neruda, R.; Kárný, M.), s. 145-148. ISBN 3-211-83651-9.
[ICANNGA'2001 /5./. Praha (CZ), 22.04.2001-25.04.2001]
R&D Projects: GA ČR GV102/96/K087; GA AV ČR IAA2030801
Institutional research plan: AV0Z1030915
Keywords : text-to-speech processing * synthetic speech * artificial neural network * training parametr for ANN
Subject RIV: BD - Theory of Information

Our effort is to minimize the difference between the synthetic speech and the natural speech. A special functional block is included into the synthesizer for prosody control. A multilayer artificial neural network is used. The number of input training parameters for ANN training must be generally kept as small as possible because of the optimal generalization ability. An original method for the determination of the most important features for the training of ANN for prosody control is described.
Permanent Link: http://hdl.handle.net/11104/0124463