do 2D NLMS adaptive channel estimation of Matlab simulation program

语音文件加窗后比较去零的范围，以及它们的比较图。-After the voice file comparison window to increase the scope of zero, and their comparison.

生成《matlab扩展编程》中对于HMM训练的samples.mat程序，自己编写-Generation ‘matlab programming expansion’ in the HMM training samples.mat procedures, written by myself

这是一个xmms功能非常强大的插件，能有卡拉OK功能，还有混响、减噪、EQ等等功能，其中混响的效果可达到一定专业水准！强烈向站长推荐！-xmms This is a very powerful plug-in, there will be karaoke function, there is reverberation, noise abatement, EQ functions, and so on, the effect of reverberation can reach a certain professional standards! Strongly recommended to the station!

自己写的关于ica语音分离的，性能不怎么样，但还是可以运行-A program for ICA speech signal separating created by myself. Although the function is not excellent, it can run.

基于LABVIEW的语音识别程序子模块，完成识别、滤波、录音、提取特征等功能

SUN G.721 sourcecode

基于GMM的概率神经网络PNN具有良好的泛化能力，快速的学习能力，易于在线更新,并具有统计学的贝叶斯估计理论基础，已成为一种解决像说话人识别、文字识别、医疗图像识别、卫星云图识别等许多实际困难分类问题的很有效的工具。而且PNN不但具有GMM的大部分优点，还具有许多GMM没有的优点，如强鲁棒性，需要更少的训练语料，可以和其他网络其他理论无缝整合等。-GMM based probabilistic neural network PNN good generalization ability, the ability to learn fast, easy online updates, and with the Bayesian statistical theory based on estimates, and has become a solution as speaker recognition, text recognition, medical image recognition, satellite images and other real recognition when difficulties classification of very effective tool. But GMM PNN is not only the most advantages, but also has many advantages GMM not as strong robustness, require less training corpus, and other networks to other theories, such as seamless integration.

hss speaker system for direct target voice transmission

HMM_隐马尔可夫过程的介绍，希望对初学者有用-HMM_ Hidden Markov process, the hope that useful for beginners

lpc10-15为美军2400bps语音压缩标准的C语音源代码。-lpc10-15 for the U.S. 2400 voice compression standard C source code voice.