For the length of feature vectors of speech samples is different, direct cutting and Dynamic Time Warping (DTW) regulation, are put forward to solve the problem.
提出了直接截取和DT W规正两种方法来解决语音样本特征向量长度不一致的问题。
2
Both analysis and experiments show that the recognition time is proportional to the square of the length of the feature vector, if dynamic time warping (DTW) algorithm is employed in recognition.