estimation_image

HMM连续语音识别(Continuous Speech Recognition)

WM8731语音芯片的驱动程序，完成录音和播放功能。主控制芯片为STM32RCT6。测试通过。(WM8731 voice chip driver to complete the recording and playback functions. The main control chip is STM32RCT6. Test pass.)

基于LMS算法的变换域的源程序，里面给出了学习曲线，迭代后的输出曲线(LMS algorithm based on the Domain source, which is a learning curve, the output iterative curve)

基于谱减法的语音降噪处理，里面有声音文件，可以直接运行观察效果(Based on the spectral subtraction speech denoising processing, there are sound files, can be run directly observed effect)

基于非线性主分量分析(PAST)的 自适应盲信号分离 (Based on nonlinear principal component analysis (PAST) Adaptive Blind Signal Separation)

分段信噪比测试，主要是用于语音增强，语音编码后的测试使用(segment SNR test, primarily for speech enhancement, speech coding after testing)

能有效实现语音合并，将两个相同采样率和相同精度的语音文件合并成为一个相同采样率和相同精度的语音文件。(Can effectively realize the merger of voice, and the two the same sampling rate and the same precision audio files into a different sampling rate and the same accuracy of voice files.)

应用LBG算法对语音信号进行矢量量化。本压缩包共有两个主文件，training.m对训练数据进行处理并得到四个初始码本，quantizing.m对待量化数据进行矢量量化。其余为自编功能函数。(application of voice signal VQ. The compressed into two main documents, training.m right training data to be processed and four of the initial code, quantizing.m treat quantitative data for vector quantization. The remainder self-function.)

一个完整的语音端点检测MATLAB程序序，包含有完整的文件，通过调试，正确可用 (An integrated voice activity detection MATLAB program, includinng complete documentation, through testing, can be used correctly )

code for voice analysis final code for sapeech recognition