资 源 简 介
Levison-Durbin 语音信号处理中的线性预测编码LPC 理论、格型滤波器以及求解现行预
测方程的算法，可以实现对语音信号重要元素的分析、合成甚至识别。
基于现有的实验平台，我们可以利用 Matlab 函数来获得几个固定语音元素（如元音）
的模型系数，LPC 得到的系数组成 IIR 滤波器。利用冲击脉冲
序列作为输入，我们就可以得到原来的语音。这是一种简单的语音合成功能。-Levison-Durbin speech signal processing in linear predictive coding LPC theory, lattice filters, as well as the current prediction equation solving algorithm, can achieve an important element of the speech signal analysis, synthesis or recognition. Based on the existing experimental platform, we can use Matlab function to obtain the number of fixed-voice elements (such as vowels) model coefficients, LPC coefficients are the composition of IIR filters. Shock pulse sequence used as input, we can get the original voice. This is a simple voice synthesis.
