TDOA_&TEST

实现了基于特定话语的聚类LGB和VQ的说话人识别(Speaker Recognition Based on Clustering LGB and VQ)

目前2.4kbps的混合激励线性预测 (MELP) 语音编码方法已经被确定为美国新的联邦语音编码标准.本文提出了一种改进的MELP语音编码方法,利用滤波器相似度和基于LPC系数分类的矢量量化技术,可以把MELP的码率降到1.7kbps以下,仍有较好的合成语音质量.(An Improved MELP Speech Coder)

espeak语音生成库，内附有移植使用文档《espeak移植使用》(espeak speech generation library, containing transplanted using the document espeak transplant use )

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!)

用于说话人识别（声纹识别）中训练过程和识别过程的高斯混合模型程序(GMM model for the training process or test process of speaker identification)

实现多个人的说话人识别，基于gmm模型，用mfcc参数训练(voice conversion based on gmm)

这是一个xmms的3D音效插件。算法简单，效率很高，如果使用音箱来听效果会更好！(xmms This is a 3D audio plugin. Algorithm is simple, highly efficient, if used to listen to speakers results will be better!)

FastICA算法，用于信号的独立分量分析，在ICA的基础上加快了收敛速度，有更高的效率！并且增加了图象界面，使用方便！(FastICA algorithm, the signal for an independent component analysis, at the ICA on the basis of accelerating the convergence rate, a more efficient! And to increase the image user interface!)

最经典的基于ICA实现的语音信号的采集、随机混合，再通过盲分离将混合后的语音信号分离(The most classical ICA-based speech signal acquisition, random mixing, and then the mixed speech signal is separated by blind separation.)

simple voice recognition program