Extract-Sound-Feature

辅助计算SDR 评测语音质量参数 评估语音质量的好坏(caculate SDR)

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).)

采样小波包分解语音信号，分解为3层，并求出分解系数(Speech signal using wavelet packet decomposition，decomposing 6 level，and attain decomposition coefficient)

Edinburgh Speech Tools for speech synthesis

说话人识别GMM模型训练MATLAB程序，需要和voicebox配合使用，使用voicebox提取MFCC和k均值分类，这是GMM说话人识别模型的训练程序。(Speaker Recognition GMM model training MATLAB program used in conjunction with and voicebox, voicebox extract MFCC and k-means classification, which is a GMM speaker recognition model training program.)

一个谱相减程序（IN MATLAB）， 一个谱相减程序（IN MATLAB），(a spectral subtraction procedure (IN MATLAB), a spectral subtraction procedure (IN MATLAB).)

LMS多麦克风语音降噪的主程序是lmsspdn.m(Multi-microphone noise reduction LMS voice is the main program lmsspdn.m)

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

MATLAB仿真LMS算法完整源代码，仿真结果分析，及仿真图示(MATLAB simulation of LMS algorithm complete source code, simulation results analysis, and simulation icon)

code for voice analysis final code for sapeech recognition