A Concatenative Turkish Text-To-Speech System : IMPLEMENTATION AND EVALUATION

This study is intended to build the framework of a concatenative TTS (Text to Speech) system for Turkish. The system is based on concatenative, unit selection approach. It contains two different speech databases comprised of units which are directly recorded and cut from a continuous speech. Some digital signal features such as zero crossing rate and energy of speech have been used. PSOLA (Pitch Synchronous Overlap and Add) algorithm has been used for smoothing while concatenating the units. Some subjective tests are used to measure the success. The quality of the synthesized speech is measured depending on intelligibility, the ability to be understood, and naturallness, the closeness to human speech. Mean Opinion Score (MOS) has been used for naturalness and Diagnostic Rhyme Test (DRT) and Comprehension Test (CT) have been applied for intelligibility. Although the system uses simple techniques, it provides promising results, since the method is very well suited for Turkish language structure. The study should be especially useful to professionals and academicians in speech processing field for Turkish like languages.