基于神经网络的说话人识别

1. 按键配置: IOA8 : 录音 // IOA9 : 停止 // IOA10: 播放 // 2. 使用IOA0~IOA7作为数据IO 0~7, IOB0~IOB5 作为控制信号线-1. Button to configure: IOA8: sound recording// IOA9: stop// IOA10: Play// 2. IOA0 ~ IOA7 use as a data IO 0 ~ 7, IOB0 ~ IOB5 as a control signal line

HTK3.4（HTK最新版)源码 + 可执行文件 + HTK book + 其他少量HTK资料（含中文）-HTK3.4 (HTK latest version) source+ Executable file+ HTK book+ Other HTK a small amount of information (including Chinese)

在有背境噪音的条件下，进行语音端点检测，并取得很好的效果-Have background in the noise conditions, for voice activity detection and achieved good results

SUN G.721 sourcecode

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

可以讲mulaw语音数据转换成线性pcm数据-can speak mulaw voice data converted into linear data pcm

此方案基于全极点语音产生模型，编码过程简述如下：首先通过线性预测分析方法提取声道滤波器参数；其次通过合成分析法确定最佳激励矢量；最后将滤波器参数和最佳激励矢量进行编码传输。-This program is based on all-pole speech production model, the encoding process summarized as follows: First, by linear prediction analysis to extract channel filter parameters followed by analysis by synthesis method to determine the best incentive vector Finally, the filter parameters and the best incentive vector encoded transmission.

用于语音识别，基于HMM模型，用C++语言编写。可用连续语音识别-It is based on HMM Model and developed with C++ which could be used to continuous speech recognition.

轻松做出类似Foxmail MSN的界面的软件-easily make similar MSN Live interface software

自编的联合对角化程序，简单，清晰，并有痃详细的注解，带有两个波形文件，可直接显示结果。-jointly staged by the diagonalization procedure, a simple, clear and detailed explanation feast, with two waveform documents, The results can be shown directly.