资源列表
matlab
- 语音信号的短时分析,主要包括:分帧、短时能量、短时平均幅度、短时过零率、短时自相关函数、短时幅度差、倒谱、复倒谱、lpc系数、lpc谱估计等(The short-time analysis of speech signal mainly includes: frame, short-time energy, short time average amplitude, short-time zero crossing rate, short-time autocorrelation functio
tensorflow_speech_recognition_demo-master
- 此程序为语音识别深度学习程序,使用python程序编写。(This program is written for the speech recognition depth learning program, using the python program.)
最终代码说话人识别
- 实现了基于特定话语的聚类LGB和VQ的说话人识别(Speaker Recognition Based on Clustering LGB and VQ)
hmm
- 实现mffc求解 并且可利用hmm进行孤立词识别
mel
- 语音信号处理中的梅尔断点检测,多次实现,原理相同,差别比较小。
dtw
- 基于MATLAB的语音识别系统,dtw - DTW算法演示程序 mfcc.m - MFCC参数计算程序 dtw.m - 基本的DTW算法 dtw2.m - 优化的DTW算法 testdtw.m - DTW算法测试程序 vad.m - 端点检测程序 -Speech Recognition System Based on MATLAB
Speech_signal_short_time_analysis
- 语音信号的短时分析,主要包括:分帧、短时能量、短时平均幅度、短时过零率、短时自相关函数、短时幅度差、倒谱、复倒谱、lpc系数、lpc谱估计等 绝对保证质量,是保研后导师布置的一些基础程序
binary_io
- this the second phase of hmm code-this is the second phase of hmm code
隐含层为mexihat 输出为sigmal的wav-sigal
- 用三层小波神经网络实现的与文本无关说话人识别。(识别部分)。输入的是语音特征,输出的是识别结果。训练用的语音特征要事先提取出。-with three wavelet neural network has nothing to do with the text of Speaker Recognition. (Recognition). The admission of voice features, the output is the result of recognition. Trainin
MATLAB-YUYINSHIBIE
- 这是语音识别的几个程序,包括语音的特征提取、端点检测的程序。-This is the number of speech recognition procedures, including the voice feature extraction, endpoint detection procedure.
matlab_reduce_noise
- matlab去除50hz噪声。 我用电脑录了一段声音,里面有50hz的周期噪声(因为受交流电干扰)。而我自己的声音频率最低是90hz。我使用了一个10阶butterworth高通滤波器,边带是70hz(介于50跟90之间)。 问题是,这不能直接用。因为声音文件的采样率是22k,70相对于22k来说太小了。所以我得先把我的声音欠采样,然后再滤波,然后再插值。
Cepstrum
- 这是语音信号处理中求倒谱的程序,我自己编的,很简单,实现了一段序列求倒谱的功能-This is the voice signal processing for cepstrum procedures, I developed, is very simple. achieve a sequence for the functional cepstrum