Collection and analysis of data for evaluation of concatenation cost functions

Abstract

This paper describes the collection and analysis of data, which are planned to be used for the evaluation and development of concatenation cost functions for unit selection based TTS systems. Data, collected via listening tests following the recommendations given in [1], were analyzed in a variety of ways to identify and possibly exclude "malicious" listeners as well as to demonstrate their sufficient "richness" for the aimed utilization. This study was limited to five Czech vowels as these sounds are characterized by being highly energetic and having rich spectral content, which induces complexity and wide range of possible discontinuities at concatenation points. © 2010 Springer-Verlag Berlin Heidelberg.