A methodology based on wavelet packet for speaker transform recognition

A novel method based on wavelet packet transform (WPT) for speaker verification and identification is proposed in this paper. In this method, WPT, for its advantage in feature extraction in high frequency, is adopted in speech signal feature extraction. After the features are extracted, the artificial neural network (ANN) is introduced as the classifier to verify and identify the speakers. The experimental results validate the effectiveness of the proposed method.