Speech Synthesis of Dissimilar Languages Using their Phonetic Superset

Chungurski, Slavco and Kraljevski, Ivan and Stojanovic, Igor and Prevalla, Blerta (2010) Speech Synthesis of Dissimilar Languages Using their Phonetic Superset. In: Proceedings of 8th Conference, Digital Speech and Image Processing, 16-18 Dec 2010, Iriski Venac, Serbia.

Text Conf17_DOGS2010_IgorStojanovic_2.pdf Download (460kB)

Abstract

The two most common languages in Macedonia are Macedonian and Albanian language. The main idea of this paper is to upgrade the phonetic inventory of Macedonian language to its superset – phonetic inventory of the Albanian language, and use it for speech synthesis in both languages.
The structure of general speech synthesis system for the Macedonian language is used [1] for its upgrade to speech synthesis system, which will produce Albanian language also. The modification of the system is described and there are details about the specific features of the languages important for the speech synthesis, like sound and vowel system and prosodic elements.
Because concatenative speech synthesizers are based on a collection of speech segments, the creation of common speech corpus for these languages is described.

Item Type:	Conference or Workshop Item (Paper)
Subjects:	Natural sciences > Computer and information sciences
Divisions:	Faculty of Computer Science
Depositing User:	Igor Stojanovik
Date Deposited:	14 Nov 2012 12:56
Last Modified:	14 Nov 2012 12:56
URI:	https://eprints.ugd.edu.mk/id/eprint/853

Actions (login required)

View Item