Fourth International Workshop on Spoken Language Technologies for Under-Resourced Languages (SLTU-2014)

St. Petersburg, Russia
May 14-16, 2014

On Mirandese Language Resources for Text-to-Speech

José Pedro Ferreira (2), Cristiano Chesi (1,3,4), Hyongsil Cho (1,3), Daan Baldewijns (1), Daniela Braga (1,3), Miguel Dias (1,3)

(1) Microsoft Language Development Center, Portugal.
(2) Instituto de Linguística Teórica e Computacional. Portugal
(3) ISCTE-IUL University Institute of Lisbon, Portugal
(4) IUSS, Istituto Universitario di Studi Superiori, Pavia, Italy

This paper aims at describing the major components of the first Text-to-Speech (TTS) system ever built for Mirandese, [1] a minority language spoken in the Northeast of Portugal. Both language resources development (corpus, textnormalization rules, annotated lexicon, phone sets and recordings) and the TTS (Statistical Parameter Synthesis) system are documented here.

Reference

  1. J. P. Ferreira, C. Chesi, D. Baldewijns, D. Braga, M.S. Dias 2012. The First Mirandese Text-to-Speech System. Proceedings of ELE2013 Conference

Index Terms: Mirandese, text-to-speech, language resources

Full Paper

Bibliographic reference.  Ferreira, José Pedro / Chesi, Cristiano / Cho, Hyongsil / Baldewijns, Daan / Braga, Daniela / Dias, Miguel (2014): "On Mirandese language resources for text-to-speech", In SLTU-2014, 87-91.