微软的语音识别资料

首先对语音信号的预处理和特征提取问题进行了讨论,提取了三个有效的语音识别特征参数—LPC系数、LPC倒谱系数和Mel频率倒谱系数(MFCC);其次着熏讨论了语音识别中常用的两种识别方法—基于模板匹配的DTW和基于统计膜型的HMM在语音识别中的应用,分别讨论了各自的模型训练和识别的方 法,对HMM在实际应用中的一些具体问题也进行了讨论;最后对常规HMM提出了一些改进方法,包括显式状态驻留和随机分段模型(SM),重点讨论了SM在 语音识别中的训练和识别方法,并通过一个孤立数字语音识别系统验证了其比常规HMM更能描述语音信号的特征。

录音部分是参照vckbase的录音api代码,结合了一小段socket(TCP)就可以用来单方说话录音了,程序分两部分一部分是(录音机+网络发送代码),一部分是(接收数据+播放波形音乐代码),由于程序只是为了自己用,很多地方没有注意错误的识别,结构也比较乱,本程序当初最头疼的就是控制损耗内存,结果用了双缓存来存储波形数据来交替的存储/清除. -part of the recording light vckbase recording api code combination of a small socket (TCP) can be used for recording a unilateral words, part of the procedure is divided into two parts (recorders network code), part of the (receiving data waveform music broadcast code), as a result of procedures only for their own use. many places no attention to the wrong identification, the structure was chaotic, most of the first procedure is a headache for the control of memory loss, Results of the double-caching to store waveform data to the turn of the storage/removals.

语音识别中的通道自相关函数，matlab版。自己编的，通过测试-Speech Recognition access autocorrelation function, Matlab version. His series, testing

提供了segmantal SNR及IS的计算代码，用于语音增强后评价用。-Provided segmantal SNR and IS calculation code for speech enhancement with post-evaluation.

语音信号处理工具包，是日本的一个大学开发的，对于HMM模型的建立很有帮助。-A tool box of speech processing, exploited by a Japanese university. It s quite helpful for setting up a HMM model.

基于zigbee协议的，无线音频传输应用，可以实现无线耳机无线音响-Zigbee-based protocol, wireless audio transmission applications, you can realize wireless headset wireless audio

Linux下的CELP语言编码，压缩语音质量极好。支持2K到32K的压缩。实现部分使用C语言编写。可用做Linux下的动态库，有例子。-under Linux CELP language coding, compression excellent speech quality. Support 2 K to 32 K compression. Part of the use of C language. Linux can be done under the DLL, is an example.

mp3算法并讲解mp3编码-mp3 algorithm on the MP3 encoder and

应用背景他是一个程序 ；VoIP ..是一个很好的项目 ；VoIP关键技术VoIP  ；（H323）源，G.723.1 G729 G728 G711音频算法的实现