Novel Estimation Method for Superpositional Intonation Model: From Text to Form

TORRES HM..; GURLEKIAN JA..

IEEE TRANSACTIONS ON AUDIO SPEECH AND LANGUAGE PROCESSING

IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC

Lugar: New York; Año: 2015

1558-7916

Fujisaki?s intonation model parameterizes the F0?s contour efficiently and giving its strong physiological basis has been successfully tested in different languages. One problem thathas not been fully addressed is the extraction of the model?s parameters, i.e., given a sentence, which of the model?s parameter values best describe its intonation. Most of the proposed methodsstrive to optimize the parameters so as to obtain the best fit for the F0 contour globally. In this paper we propose to use text information from the sentence as the main guide or reference foradjusting the parameters. We present a method that defines a set of rules to fix and optimize the model?s parameters. Optimization never loses sight of the events of the text?s structure that arouseit. When text information is not enough, the algorithm predicts parameters from F0 contour and tie it to the text. The process of parameter estimation can be seen as a way to go from textinformation to the F0 contour. The parameter optimization is carried out to fit the F0 contour locally. Our novel approach can be implemented manually or automatically. We present examplesof manual implementation and the quantitative results of the automatic one. Tested on three corpora in Spanish, English and German, our automatic method show an performance of 34%better than other tested methods.