Roles of the Average Voice in Speaker-adaptive HMM-based Speech Synthesis

Abstract

In speaker-adaptive HMM-based speech synthesis, there are a few speakers whose synthetic speech sounds worse than that of other speakers, despite having the same amount of adaptation data from within the same corpus. This paper investigates these fluctuations in quality and found that as mel-cepstral distance from the average voice becomes larger, the MOS scores generally become worse. Although the negative correlation obtained is not strong enough, this helps us improve the training and adaptation strategies for average voice models. Furthermore we remark that this correlation is strongly linked to “vocal attractiveness.”