Speech parameter generation from HMM using dynamic features.

Keiichi Tokuda, Takao Kobayashi, Satoshi Imai

Browse the full ICASSP paper archive.