Compared with its forerunners, the proposed method uses pronunciation knowledge of homophones in a more complex way.
確定! 回上一頁