Duration modeling for text to speech synthesis system using festival speech engine developed for Malayalam language

Author

Rajan, Bindhu K. ; Rijoy, V. ; Gopinath, Deepa P. ; George, Nimmy

fYear

2015

fDate

19-20 March 2015

Firstpage

1

Lastpage

5

Abstract

This paper describes duration modeling in Text To Speech Synthesis (TTS) for Malayalam language using open source Festival TTS engine. Classification and Regression Tree (CART) based data-driven phoneme duration modeling is presented. A number of features are extracted for predicting the duration of phonemes. Objective evaluation test was conducted to evaluate the intelligibility of the synthesized speech by root mean squared error (RMSE) and correlation between actual and predicted durations. The objective evaluation of the model gave an RMSE of 0.1188 and a correlation of 0.9918.

Keywords

feature extraction; mean square error methods; natural language processing; regression analysis; signal classification; speech processing; speech synthesis; trees (mathematics); CART; Malayalam language; RMSE; actual durations; classification-and-regression tree based data-driven phoneme duration modeling; feature extraction; objective evaluation test; open source festival TTS speech engine; predicted durations; root mean squared error; text-to-speech synthesis system; Computational modeling; Correlation; Feature extraction; Hidden Markov models; Integrated circuit modeling; Speech; Speech synthesis; CART; Festival; TTS synthesis; features;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuit, Power and Computing Technologies (ICCPCT), 2015 International Conference on

Conference_Location

Nagercoil

Type

conf

DOI

10.1109/ICCPCT.2015.7159332

Filename

7159332