coherence_MVDR_jacob-benesty

实现声源定位功能，利用两个麦克风的语音数据获得原始声源的位置!(Achieve sound localization function, using two microphones for voice access to the original source data location!)

鲁棒数字音频盲水印水印置乱算法实现基于阿诺德变换的离散小波变换，离散余弦变换(Robust Blind Digital Audio Watermarking Implementation Based on Watermark Scrambling Algorithm-Arnold Transform Discrete Wavelet Transform,Discrete Cosine Transform,Erro Correcting Code)

科大讯飞关于语音识别开发SDK库，支持语音转汉字功能及根据读汉字功能。有兴趣的可以好好研究研究(IFLYTEK speech recognition development SDK library are interested in a good studies)

端点检测算法采用现阶段比较精确的时频方差检测法，模式识别和模板匹配采用的是改进后的DTW算法（主要是限定了DTW的搜索路径，进一步精确了DTW的平行四边形的形状，进一步的减少匹配中的参数存贮量以及多余的搜索路径）。(Endpoint detection algorithm uses more precise at this stage variance time-frequency detection, pattern recognition and template matching is based on the improved DTW algorithm (DTW is mainly limited to the search path, and further precision of the DTW parallelogram shape, and further reduction of matching the parameters in the storage volume as well as the extra search path).)

mfcc参数提取过程，具体内容可以参看参考(MFCC parameter extraction process)

进行语音合成和分析，最基本的数据来源和可靠合成基础(Speech synthesis and analysis, basic data sources and reliable synthetic base)

端点检测程序 多频带，熵 谱熵 teger能量 希望大家好好学习好好利用(endpoint detection procedures multiband, entropy spectral entropy teger energy hope that we learn to make good use)

hmm文件时运用HMM算法实现噪声环境下语音识别的。其中vad.m是端点检测程序；mfcc.m是计算MFCC参数的程序；pdf.m函数是计算给定观察向量对该高斯概率密度函数的输出概率；mixture.m是计算观察向量对于某个HMM状态的输出概率，也就是观察向量对该状态的若干高斯混合元的输出概率的线性组合；getparam.m函数是计算前向概率、后向概率、标定系数等参数；viterbi.m是实现Viterbi算法；baum.m是实现Baum-Welch算法；inithmm.m是初始化参数；train.m是训练程序；main.m是训练程序的脚本文件；recog.m是识别程序。(hmm HMM algorithm file using speech recognition in noisy environments. Which is the endpoint detection process vad.m mfcc.m procedure is to calculate the MFCC parameters pdf.m function is calculated for a given observation vector of the Gaussian probability density function of output probability mixture.m is to calculate the observation vector for a HMM state output probability of observation vector is the number of Gaussian mixture per state output probability of the linear combination getparam.m before the calculation of the probability function, backward probability, calibration coefficients and other parameters viterbi.m is Viterbi algorithm implementation baum.m Baum-Welch algorithm to achieve inithmm.m is the initialization parameters train.m is the training program main.m training program is a script file recog.m is to identify procedures.)

关于lms经典算法汇总，包括lms,nlms,LMS_Equalizer,Basic_LMS,LMS toolbox等等。(classical algorithm for matrix lms, including lms, nlms. LMS_Equalizer, Basic_LMS, LMS toolbox, and so on.)

简单语音识别源码是一个调用讯飞语音识别SDK的例子源码是一个最纯净的Demo比较容易看懂。实现的是点击按钮开始语音监听，手机需要联网，2/3G的均可，识别后会以Toast的方式提示出来，识别率还是比较高的。之前自己是用别人的Demo总是不成功，是因为，讯飞的SDK想使用是需要APPID的，可能过太长时间久会注销，还有，未通过审核的APPID每天只能试用500次，如果太多人试用，就是达到上限导致无法使用。(A simple speech recognition source code is a call Ifly speech recognition SDK example source code is a demo of the purest easier to understand. The realization of the click button to start the voice listening, mobile phone need to networking, 2/3G can, after the recognition will be prompted by Toast way out, the recognition rate is still relatively high. Before he is with someone else s demo is not always successful, is because the Xunfei SDK to use is appid, may be too long for a long time will be canceled, and not through the audit appid every day can only be tried out 500 times, if too many people tried is reaching the limit to use.)