Recent improvements of Probability Based Prosody Models for Unit Selection in concatenative Text-to-Speech

The work presented in this paper is subsequent to the paper ldquoProbability Based Prosody Model for Unit Selectionrdquo which was published in ICASSP´2004. In the improved probability prosody model for corpus based concatenative Text-to-Speech (TTS), likelihood is replaced with posterior probability in the cost functions which conduct the following step, unit selection. Objective and subjective experiments show that posterior probability has obvious advantages over likelihood on robustness, flexibility and overall quality.