微软研究出来的一个基频提取算法，简单有效

微软speechsdk的语音控制开发例子-Microsoft speechsdk example of voice control developed

用于语音识别的离散隐马尔可夫模型程序（非特定人，多个观测序列），已调试通过。-used discrete speech recognition Hidden Markov Model (non-specific, a number of observation sequence), has been through testing.

语音识别中的MFCC参数提取的java源码和FFT的java源码-Speech Recognition MFCC parameter extraction of the java source and FFT of the java source code

加入短时能零积端点检测技术的幅值谱减法，噪声谱的估计根据端点检测的结果进行加权平均，时时更新，更好的去除噪音，并将去噪后的结果写成wav文件-Adding short-term endpoint detection can be zero plot of the amplitude of spectral subtraction techniques, the estimated noise spectrum endpoint detection in accordance with the results of the weighted average, up-to-date, better to remove noise, and after the results of denoising wav file

经典LMS算法极其示例,希望有人能让我分享个多输入的LMS算法-LMS algorithm is extremely classic example of the hope that someone will allow me to share more than the LMS algorithm input

语音识别专辑二- Speech recognition special edition two

用于汉语语调识别的程序，可以识别我们说话中的四个声调。已通过测试-Chinese tone recognition for the program, has been tested

基于Matlab编写的语音端点检测程序，比较基础，希望可以帮到需要的人-prepared based on Matlab voice endpoint detection procedures, a comparable basis, the desire to help people in need

本程序实现用谱减法消除语音中的噪声，其中语音是纯净的，噪声是后加上去的，滤除效果良好-This procedure using spectral subtraction to eliminate the noise in the voice, which voice is pure, after noise is added, and filter out the effect of good

基于microsoft speech sdk的语音识别代码程序-Microsoft speech sdk-based speech recognition procedures code