Text-to-Speech Systems for Romanian Language

Abstract

This paper presents the text-to-speech systems developed by our team at the Technical University of Cluj-Napoca. The speech-synthesis methods we implemented are the formantic, the phoneme-and diphoneme-based, the Multipulse Excited LPC, the Nonlinear Springing Method, the Repeat-Remove PSOLA and the HMM-TTS called HTS.