Abstract
"Singing" is one of the most important and popular artistic human activities. The purpose of this research is to develop a mechanical Singer Robot with artificial larynx. We have already developed a singer robot that can generate Japanese 5 vowels [a], [i], [u], [e], and [o], and nasal plosive consonants. The robot has a mechanism to change the voice pitch by manipulating artificial vocal cord. In this paper, new mechanisms for generating other consonants, such as plosive, stop, affricate, fricative, nasal consonant and semivowel are presented. Experimental results show the validity of those mechanisms.
