... without reference sentences based on deep neural network acoustic models ... of an utterance (the reference sentence) when evaluating the utterance, ...
確定! 回上一頁