No abstract

The syllable is a speech synthesis unit that is successfully used in high quality speech synthesis systems. In support of a syllable based speech synthesis system, we consider necessary the realization of an algorithm for automatic syllabication. In this paper we present an algorithm for automatic syllabication. The algorithm starts by making the phonetic transcription of the word using a previously developed method which employs a set of phonetic rules represented through a XML scheme. Then it applies a set of rules for syllabication. The system obtained a 10.5% word error rate for syllabication and 4.79% word error rate for phonetic transcription using rules.

No abstract

No abstract

This paper presents an evaluation of 5 letter-tosound (LTS) systems for Romanian. The first is an expert system; three of them use automatic classification methods with decision trees, neural networks and support vector machines respectively and the fifth one uses pronunciation by analogy. All systems were trained and tested on the same database: a 15,517 words corpus built according to the SpeechDat specifications and a corpus consisting of the most frequent 4779 words in Romanian. The results show that decision trees and neural networks generate the best results for letter to sound conversion in Romanian.