语音识别程序

将文本按照标点符号分开的程序源码,可以把分开后的源码分开保存到文本文件中.为方便语音识别软件的制作而开发.-punctuation separate proceedings source, can be separated from the source to maintain separate text file. For the convenience of voice recognition software production and development.

NLP技术实现，对语料库进行自动统计生成分词词典，对训练集进行分词，列出所有的分词可能并计算每种可能的概率。请使用者自行加入语料库和测试集。-NLP technology to automatically Corpus Health Statistics ingredients dictionary, the training set for segmentation, list all the sub-term may calculate the probability of each possible. Requests for users to set and test sets.

非线性量化。读取wav文件，用2到12比特数量化。抽样频率8KHz。

一个简易的端点检测程序，已经实现，包括基于过零率、短时能量的双门限判决方法。-a simple endpoint detection procedures has been achieved, including on the basis of zero rate Short-term dual energy threshold judgment method.

不定长信号合并，％执行信号相加： y(n) = x1(n)+x2(n) ％调用方式：[y,n] = sigadd(x1,n1,x2,n2) ％ y：叠加后的信号序列的序列号，包括n1和n2 ％ x1： 第一个信号序列，序列号为n1 ％ x2：第二个信号序列，序列号为n2(n2可以与n1不同)-Indefinite long-Signal merger, add up to the implementation of signal: y (n) = x1 (n)+ X2 (n) call the way: [y, n] = sigadd (x1, n1, x2, n2) y: Superposition signal sequence after the serial number, including the n1 and n2 x1: the first signal sequence, the serial number for n1 x2: the second signal sequence, the serial number for n2 (n2 with n1 different)

语音信号处理中的所有基本的算法的代码，很全的。欢迎大家下载使用:)-almost all of the basic algorithm code in speech signal processing. welcome to download.

Opencore library for AMR-NB and AMR-WB

基于概率神经网络方法的识别程序,主要针对语音识别,matlab程序,值得参考.-probabilistic neural network-based method of identification procedures, voice recognition, Matlab procedures, a good reference.

提供rls的matlab实现，用RLS算法设计一个数字滤波器，实现语音增强的目标，得到清晰的语音信号-Provide rls realize the matlab using RLS algorithm to design a digital filter to achieve the goal of speech enhancement have been given clear voice signal

语音识别与合成入门5，word或压缩包中内有matlab源码-speech recognition and synthesis 5 entry, word or compressed packages within Matlab FOSS