Data-driven joint f/sub 0/ and duration modeling in text to speech conversion for Spanish | Publicación