摘要
从人类语音产生的机理出发,介绍了语音信号的特征和语音信号的语谱图,引出了语音信号的产生模型。同时讨论了在语音信号产生的模型应用中,线性预测编码方法及语音产生模型在语音合成和语音识别中的应用原理,体现了语音产生模型在语音处理技术方面的重要地位。
This paper introduced the pronunciation signal characteristic and the pronunciation signal language spectrogram, which the mechanism that pronciation produces set off from human being simultaneously. It discussed the model application in pronunciation signal produces, line predictive coding method in the pronunciation signal and the pronunciation produces the model in the pronunciation synthesis and the speech recognition application principle, manifested the important status that the pronunciation produces the model in the pronunciation processing technology aspect.
