Publication detail

Text-to-Speech System Modification by Soft Computing Methods

Jan Čermák

Original Title

Text-to-Speech System Modification by Soft Computing Methods

Type

conference paper

Language

English

Original Abstract

Machine human communication is becoming more and more popular by the public and also by commercial sector. One of the possible machine human communications is a transformation of a text to speech. Systems performing this transformation, usually with unrestricted dictionary, are called Text-to-speech (TTS) systems. Nowadays, TTS systems are utilized in web browsers, email reading application, telephone services etc. But although the TTS systems usage is very wide, they are still not often used. This is mainly caused by insufficient prosody modeling of synthesized utterance. Prosody modeling is a language dependent complex task, which has not been sufficiently solved yet. Scientists are trying to find new approaches to improve prosody modeling. One of the possible approaches is introduced in this paper.

Key words in English

TTS, fuzzy logic, neural network, speech

Authors

Jan Čermák

RIV year

2005

Released

1. 1. 2005

Publisher

Educational Sports and Recovery Complex of the TU-Sofia

Location

Sozopol

ISBN

9544385177

Book

ELECTRONICS ET 2005. The Fourteenth International Scietific and Applied Science Conference Electronics ET'2005

Pages from

25

Pages to

54

Pages count

30

BibTex

@inproceedings{BUT16143,
  author="Jan {Čermák}",
  title="Text-to-Speech System Modification by Soft Computing Methods",
  booktitle="ELECTRONICS ET 2005. The Fourteenth International Scietific and Applied Science Conference Electronics ET'2005",
  year="2005",
  pages="30",
  publisher="Educational Sports and Recovery Complex of the TU-Sofia",
  address="Sozopol",
  isbn="9544385177"
}