Prosody generation model for TTS systems - segmental durations and F0 contours with fujisaki model Livro