mfcc参数提取过程

回音消除算法matlab,采用的VIRE的双方通话检测(Matlab algorithms eliminate echoes, the two sides call the VIRE Detection)

应用LBG算法对语音信号进行矢量量化。本压缩包共有两个主文件，training.m对训练数据进行处理并得到四个初始码本，quantizing.m对待量化数据进行矢量量化。其余为自编功能函数。(application of voice signal VQ. The compressed into two main documents, training.m right training data to be processed and four of the initial code, quantizing.m treat quantitative data for vector quantization. The remainder self-function.)

用自相关函数提取语音信号基频，提取音频文件的基频等高线(Use the autocorrelation function on segments of the signal (windowsize: 100ms) and compute the fundamental frequency. Use a max_time_lag of 100ms in the autocorrelation function and a window shift of 25ms. Create a fundamental frequency vector and plot the pitch contour.)

男女变声器，语言信号处理，录音，matlab，各种滤波器，有预处理，自定义时长录音并保存，自主选择文件，自主制作界面(Male and female voice changer, speech signal processing, audio, MATLAB, various filters, pretreatment, long recording and save custom, choose file, independent production interface)

说明：  C# 对接 Java 3des md5WithRsa, 有需要的可以参考下(If you need C # to dock Java 3des md5WithRsa, you can refer to the following)

c#写的语音识别程序，识别效率还算可以。且根据识别可以给出相应的识别百分比。适合研究语音识别以及tts的参考！(c# to write the speech recognition process, identify efficiency figure. And recognition can be given according to the percentage of the corresponding identification. Tts speech recognition, as well as for research reference!)

量子粒子群算法应用。对三个水听器的位置定位，在假设水听器位置已知情况下，以及其他方法估计得到的时延，通过量子粒子群算法计算，验证量子粒子群算法的计算精度。(Quantum particle swarm optimization application. Position location of the three hydrophones in the hydrophone position assumed known case, and other methods for estimating time delay obtained by calculating the quantum PSO QPSO verify the accuracy of the algorithm.)

基于GMM模型实现语音转换,通过GMM模型对谱包络的建模，实现男女，个性化语音之间转换(Voice conversion using GMM model)

说明：  HMM的演示源程序，详细解释了HMM的过程，实现了基本的语音识别功能(HMM demonstration source program, explained the process of HMM in detail, realized the basic speech recognition function)

说明：  一个比较简单粗略的基于VQ矢量化模型的说话人识别程序(Rough a relatively simple model based on the VQ vector of speaker recognition procedures)