and reduce the enormous quantity of context combinations. To evaluate the proposed HMM structure, we construct three Mandarin speech synthesis systems.
確定! 回上一頁