Building a Finnish unit selection TTS system

Hanna Silén, Elina Helander, Konsta Koppinen, Moncef Gabbouj

Abstract: Speech synthesis based on unit selection can produce far more natural speech than conventional diphone-based methods. Unit selection based text-to-speech synthesizers have been built for many different languages. In this paper, we describe the development of TUT VOICE, the first Finnish unit selection synthesis engine for academic research. The system includes database construction, synthesis engine implementation and optimization for Finnish.