voice_analysis

语音唤醒就是在设备待机状态下，用户说出特定的语音指令（唤醒词）使设备进入工作状态或完成某一操作。设备为了实现语音唤醒功能，就需要设备实时监听，也就是实时的录音并分析有没有唤醒词来唤醒设备。评价一个语音唤醒系统的标准有三个：唤醒正确率、误报率、功耗。一般情况下，唤醒正确率越高 ，误报率也越高。好的系统就需要唤醒率高，误报率低，功耗低。(Speech wakeup is when the device is in standby mode, the user speaks a particular voice command (wake up word), so that the device enters the work state or completes an operation. In order to realize the function of speech wakeup, the device needs real-time monitoring, that is, real-time recording and analyzing whether or not wake words are used to wake up the equipment. There are three criteria for evaluating a voice wakeup system: wakeup accuracy, false positive rate, and power consumption. Generally, the higher the correct rate of arousal, the higher the false positive rate. A good system requires high wake-up rate, low false positive rate and low power consumption.)

一种关于倒谱的实现研究处理方法的总结以及展望(One kind of research on cepstrum processing method to achieve a summary and outlook)

基于LPC的语音合成，可以实现变调变速功能，附带注释便于学习。(LPC-based speech synthesis, can achieve modulation speed function, annotated easy to learn.)

用VC++实现的关于语音识别中提取MFCC程序，有界面，程序能运行成功。(Use of the vc of voice recognition mfcc program, the interface, the application can run. )

用vq算法进行说话人识别，能达到良好的效果(Using VQ speaker recognition algorithm can achieve good results)

传统语音增强算法，谱减法和MCRC算法，matlab源代码。(Traditional speech enhancement algorithm, spectral subtraction and MCRC algorithm, matlab source code.)

使用高斯混合模型（GMM），对声音信号进行识别，包括对特征参数的采集、训练和识别过程(Using the Gaussian mixture model (GMM)to realize the recognition of the sound signal, including the acquisition of the characteristic parameters,the training and the recognition process)

基于最小均方误差的助听器语音去噪算法, 用于以最大程度消除助听器接受的语音杂音.(Speech denoising algorithm based on the minimum mean square error of a hearing aid for the hearing aid acceptable voice murmur eliminated to the greatest extent.)

简单语音识别源码是一个调用讯飞语音识别SDK的例子源码是一个最纯净的Demo比较容易看懂。实现的是点击按钮开始语音监听，手机需要联网，2/3G的均可，识别后会以Toast的方式提示出来，识别率还是比较高的。之前自己是用别人的Demo总是不成功，是因为，讯飞的SDK想使用是需要APPID的，可能过太长时间久会注销，还有，未通过审核的APPID每天只能试用500次，如果太多人试用，就是达到上限导致无法使用。(A simple speech recognition source code is a call Ifly speech recognition SDK example source code is a demo of the purest easier to understand. The realization of the click button to start the voice listening, mobile phone need to networking, 2/3G can, after the recognition will be prompted by Toast way out, the recognition rate is still relatively high. Before he is with someone else s demo is not always successful, is because the Xunfei SDK to use is appid, may be too long for a long time will be canceled, and not through the audit appid every day can only be tried out 500 times, if too many people tried is reaching the limit to use.)

对一段语音进行变速变调等多种处理的MATLAB样例程序。(Variable-speed modulation of a variety of voice processing sample MATLAB program.)