Aiming at the improvement of the quality of synthetic speech generated by our native TTS ARTIC, we adopted the unit selection method. Our unit selection module is driven by prosody described solely by high-level symbolic features which are linked to the prosody of synthesized phrases through the phenomena of prosodic synonymy and homonymy. It was confirmed that such an approach not only generates speech with high naturalness but also keeps the richness of prosody. Our first version of this approach significantly increased the quality of the output speech, which was assessed by listeners as very close to natural.

The concept of prosodic synonymy and homonymy is, therefore, further extended and formally described in this paper, and its importance to the unit selection treatment is demonstrated. In addition, the difference of this concept from the concepts most frequently used is shown. Moreover, the first experiment following the formal definition of the problem presented in this paper has been carried out, proving that the whole concept is feasible.