Delphi PSK无线通信，基于COM的原代码的声卡…

研究生课程：语音识别(吴善培)- Graduate student curriculum: Speech recognition (Wu Shanpei)

系统辨识与自适应控制MATLAB仿真CH5,好东西大家分享-System Identification and Adaptive Control of MATLAB simulation of CH5, a good thing to share with you

此方案基于全极点语音产生模型，编码过程简述如下：首先通过线性预测分析方法提取声道滤波器参数；其次通过合成分析法确定最佳激励矢量；最后将滤波器参数和最佳激励矢量进行编码传输。-This program is based on all-pole speech production model, the encoding process summarized as follows: First, by linear prediction analysis to extract channel filter parameters followed by analysis by synthesis method to determine the best incentive vector Finally, the filter parameters and the best incentive vector encoded transmission.

整数的语音合成 可以设置各个镀银的时延，适合初学者使用-Integer speech synthesis can set the delay all silver, for beginners

在C/C++实现中。回放录音程序，适合初学者用

利用倒谱求出语音信号频谱包络和基音频率。-Cepstrum was calculated for the spectrum envelope of speech signal and the pitch frequency.

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.

此程序提供一个类来计算快速傅里叶变换(FFT) 计算离散傅里叶变换(DFT) 和它的逆矩阵。

26104-610.zip为欧洲G3的语音压缩标准gsmamr的C语言浮点算法的标准源程序。-26104- 610.zip for the European G3 voice compression standard C language gsmamr Floating Point Algorithm standard source.

Hammerstein系统识别，其状态空间子系统包括形成的Hammerstein系统的线性部分。