Personalized natural speech synthesis based on retrieval of pitch patterns using hierarchical Fujisaki model | IEEE Conference Publication | IEEE Xplore