【基础教程】基于matlab工具voicebox函数中文说明【含Matlab源码 032期】

一、简介

1 音频文件输入或输出

readwav       - 读取WAV文件
writewav      - 写WAV文件
readhtk       - 读 HTK waveform文件
writehtk      - 写 HTK waveform 文件
readsfs       - 读 SFS文件
readsph       - 读 SPHERE/TIMIT waveform 文件
readaif       - 读 AIFF Audio Interchange file format 文件
readcnx       - 读 BT Connex database 文件
readau        - 读 AU文件(from SUN)
readflac      -读 FLAC 文件

  

 

  
1
  
2
  
3
  
4
  
5
  
6
  
7
  
8
  
9
  
10
 

2 频率尺度转换

frq2bark      - Convert Hz to the Bark frequency scale利用基本频率hz转换到Bark频率尺度
frq2cent      - Convert Hertz to cents scale利用基本频率hz转换到cents尺度
frq2erb       - Convert Hertz to erb rate scale利用基本频率hz转换到erb比例尺度
frq2mel       - Convert Hertz to mel scale利用基本频率hz转换到梅尔尺度
frq2midi      - Convert Hertz to midi scale of semitones利用基本频率hz转换到MIDI文件音高
bark2frq      - Convert the Bark frequency scale to Hz 利用Bark频率尺度转换到基本频率hz
cent2frq      - Convert cents scale to Hertz利用cents尺度转换到基本频率hz
erb2frq       - Convert erb rate scale to Hertz利用erb比尺度转换到基本频率hz
mel2frq       - Convert mel scale to Hertz利用梅尔尺度转换高基本频率hz
midi2frq      - Convert midi scale of semitones to Hertz利用midi文件音高转换到基本频率hz

  

 

  
1
  
2
  
3
  
4
  
5
  
6
  
7
  
8
  
9
  
10
 

3 傅里叶Fourier/离散余弦DCT/离散哈脱莱Hartley 变换

rfft          - FFT of real data实数的傅里叶变换
irfft         - Inverse of FFT of real data实数的反傅里叶变换
rsfft         - FFT of real symmetric data实对称数据的傅里叶变换
rdct          - DCT of real data实数的离散余弦变换
irdct         - Inverse of DCT of real data实数的反离散余弦变换
rhartley      - Hartley transform of real data实数的离散哈脱莱变换
zoomfft       - calculate the fft over a portion of the spectrum with any resolution任意分辨率的频谱傅里叶计算变换
sphrharm      - calculate forward and inverse shperical harmonic transformations正向和反向球面谐波计算变换

  

 

  
1
  
2
  
3
  
4
  
5
  
6
  
7
  
8
 

4 Probability Distributions概率分布

berk2prob     - Convert Berksons to probability利用berk转换到probability概率
gaussmix      - Fit a gaussian mixture model to data values拟合高斯混合模型的数据
gaussmixd     - Calculate marginal and conditional density distributions and perform inference边际和条件密度推挤计算
gaussmixk     - Estimate Kuleck-Leibler divergence between two GMMs两个高斯混合模型交叉熵散度估测
gaussmixg     - Calculate global mean, covariance and mode of a Gaussian mixture高斯混合的全均值，协方差，模态计算
gaussmixm     - Estimate mean and variance of GMM vector magnitude高斯混合模型向量幅度均值、方差估计
gaussmixp     - Calculates and plots full and marginal probability density from a GMM高斯混合模型边缘概率密度的计算和绘制
gaussmixt     - multiplies two GMMs together两个高斯混合模型相乘
gausprod      - Calculate the product of multiple gaussians多个高斯结果的计算
gmmlpdf       - OBSOLETE - use gaussmixp instead过时，使用gussmixp代替此函数
histndim      - N-dimensional histogram (+ plot 2-D histogram)N维直方图（+绘制二维直方图）
lognmpdf      - Prob density function of a lognormal distribution对数正态概率密度函数
maxgauss      - Calculate the mean and variance of max(x) where x is a gaussian vector一个高斯向量均值或方差的最大值计算
normcdflog    - Calculate the log of the Normal cdf without underflow没有下溢的正常CDF日志文件计算
prob2berk     - Convert probability to Berksons利用probability概率转到berk
randvec       - Generate random vectors产生随机向量
randiscr      - Generate discrete random values with prescribed probabilities生成规定概率的离散随机值
rnsubset      - Select a random subset选择的一个随机子集
randfilt      - Generate filtered random noise without transients产生无瞬变的滤波随机噪声
stdspectrum   - Generate standard audio and speech spectra生成标准音频和语音谱
usasi         - Generate USASI noise (obsolete: use stdspectrum instead)过时，用stdspectrum函数代替
v_chimv       - Approximate mean and variance of non-central chi distribution非中心分布的近似均值和方差
vonmisespdf   - Calculate the pdf of the Von Mises (circular normal) distribution计算米塞斯分布（循环正常）的pdf

  

 

  
1
  
2
  
3
  
4
  
5
  
6
  
7
  
8
  
9
  
10
  
11
  
12
  
13
  
14
  
15
  
16
  
17
  
18
  
19
  
20
  
21
  
22
  
23
 

5 Vector Distances向量距离

disteusq      - Calculate euclidean/mahanalobis distances between two sets of vectors两个向量集合的欧式距离和马氏距离
distchar      - COSH spectral distance between AR coefficient sets AR系数集之间的双曲余弦谱距离
distitar      - Itakura spectral distance between AR coefficient sets AR系数集之间的Itakura谱距离
distisar      - Itakura-Saito spectral distance between AR coefficient sets AR系数集之间的ltakura-Saito 谱距离
distchpf      - COSH spectral distance between power spectra 功率谱间的双曲余弦谱距离
distitpf      - Itakura spectral distance between power spectra 功率谱间的ltakura谱距离
distispf      - Itakura-Saito spectral distance between power spectra 功率谱间的ltakura-saito谱距离

  

 

  
1
  
2
  
3
  
4
  
5
  
6
  
7
 

6 Speech Analysis语音分析

activlev      - Calculate the active level of speech (ITU-T P.56)估算语音的活跃程度
activlevg     - Calculate the active level of speech robustly to added noise估算语音有力的加性噪声活跃程度
dypsa         - Estimate glottal closure instants from a speech waveform语音波形声门闭合时刻估计
enframe       - Divide a speech signal into frames for frame-based processing语音信号分成基于帧的分帧处理
correlogram   - calculate a 3-D correlogram三维相关图计算
ewgrpdel      - Energy-weighted group delay waveform延迟波形的能量给加权
fram2wav      - Interpolate frame-based values to a waveform波形中插入帧值
filtbankm     - Transformation matrix for a linear/mel/erb/bark-spaced filterbank from dft output 线性/梅尔/erb/bark-spaced滤波器组转换矩阵从偏流输出
fxpefac       - PEFAC pitch tracker pefac基音跟踪
fxrapt        - RAPT pitch tracker       rapt（图像？）基音跟踪
gammabank     - Calculate a bank of IIR gammatone filters     IIRgammabakn滤波器计算
importsii     - Calculate the SII importance function (ANSI S3.5-1997)SII重要函数计算
modspect      - Caluclate the modulation specrogram  调制specrogram计算
mos2pesq      - Convert MOS values to equivalent PESQ scores   MOS值等效转换到PESQ得分
overlapadd    - Reconstitute an output waveform after frame-based processing重建一个基于帧处理后的输出波形
pesq2mos      - Convert PESQ scores to equivalent MOS values  PESQ得分等效转换到MOS值
phon2sone     - Convert signal levels from phons to sones信号电平从phons转换到sones
psycdigit     - Experimental estimation of monotonic/unimodal psychometric function using TIDIGITS单调/单峰心理功能使用TIDIGITS实验估计
psycest       - Experimental estimation of monotonic psychometric function单调心理功能函数实验估计
psycestu      - Experimental estimation of unimodal psychometric function 单峰心理功能函数实验估计
psychofunc    - Psychometric functions心理功能
v_sigma       - Identify glottal closure and opening intstants from Lx or EGG waveform利用Lx或蛋波形识别声门的开闭
snrseg        - Segmental SNR and Global SNR calculation分段信噪比和全信噪比计算
sone2phon     - Convert signal levels from sones to phons信号电平sones转换到phons
soundspeed    - Returns the speed of sound in air as a function of temperature返回声音在空气的速度于温度变化的函数
spgrambw      - Spectrogram with many options声谱图的许多选项
stoi2prob     - Convert STOI intelligibility measure to probability of correct recognition标准清晰度测量转换到正确识别概率
txalign       - Align two sets of time markers两套时间标记集对齐
vadsohn       - Voice activity detector语音活动侦测器
v_ppmvu       - Calculate the PPM, VU or EBU levels of a signal计算信号的PPM、VU、EBU水平

  

 

  
1
  
2
  
3
  
4
  
5
  
6
  
7
  
8
  
9
  
10
  
11
  
12
  
13
  
14
  
15
  
16
  
17
  
18
  
19
  
20
  
21
  
22
  
23
  
24
  
25
  
26
  
27
  
28
  
29
  
30
 

7 LPC Analysis of Speech 语音线性功能控制器LPC分析

ccwarpf       - warp complex cepstrum coefficients复倒谱系数的变形
lpcauto       - LPC analysis: autocorrelation method LPC分析 自相关法
lpcbwexp      - Bandwidth expansion of LPC filter LPC滤波器的带宽扩展
lpccovar      - LPC analysis: covariance method LPC分析 协方差分析
lpcconv       - Arbitrary conversion between LPC representations LPC表示的任意转换
lpcifilt      - inverse filter a speech signal语音信号的逆滤波器
lpcrand       - create random stable filters创建随机稳定的滤波器
lpcrr2am      - Matrix with all LPC filters up to order p矩阵用LPC滤波器到p阶
lpcstable     - check for stability and force stable filters稳定滤波器的稳定和力量检查
lpc--2--      - Convert between alternative LPC representation替代LPC表示的转换

  

 

  
1
  
2
  
3
  
4
  
5
  
6
  
7
  
8
  
9
  
10
 

8 Speech Synthesis语音合成

sapisynth     - Text-to-speech synthesis of a string or matrix 字符串的文本或矩阵到语音的合成
glotros       - Rosenberg model of glottal waveform声门波形的罗森堡模型
glotlf        - Liljencrants-Fant model of glottal waveform声门波形到liljencrants-Fant模型

  

 

  
1
  
2
  
3
 

9 Speech Enhancement语音增强

estnoiseg     - Estimate the noise spectrum from noisy speech using MMSE method利用最小均方差MMSE方法从噪音中估算噪声频谱
estnoisem     - Estimate the noise spectrum from noisy speech using minimum statistics利用最小统计从噪音中估算噪声频谱
specsub       - Speech enhancement using spectral subtraction采用谱减法增强语音
ssubmmse      - Speech enhancement using MMSE estimate of spectral amplitude or log amplitude采用MMSE估计谐振幅或对数振幅增强语音
ssubmmsev     - Speech enhancement using MMSE estimate and VAD-based noise estimation利用最小均方法估计法和基于VAD的噪声估计法增强语音
specsubm      - (obsolete algorithm) Spectral subtraction 过时。谱减法
spendred      - Speech Enhancement and Dereverberation (Doire's algorithm)语音增强和混响（doir算法）

  

 

  
1
  
2
  
3
  
4
  
5
  
6
  
7
 

10 Speech Coding语音编码

lin2pcmu      - Convert linear PCM to mu-law PCM线性PCM转换到μ律PCM
pcma2lin      - Convert A-law PCM to linear PCM A律PCM转换到性PCM
pcmu2lin      - Convert mu-law PCM to linear PCM μ律PCM转换到线性PCM
lin2pcma      - Convert linear PCM to A-law PCM A律PCM转换到线性PCM
kmeanlbg      - Vector quantisation: LBG algorithm矢量量化  LBG算法
kmeanhar      - Vector quantization: K-harmonic means矢量量化 调和平均算法
potsband      - Create telephone bandwidth filter电话带宽过滤器创建
v_kmeans      - Vector quantisation: k-means algorithm矢量化 k均值聚类算法

  

 

  
1
  
2
  
3
  
4
  
5
  
6
  
7
  
8
 

11 Speech Recognition语音识别

melbankm      - Mel filterbank transformation matrix梅尔滤波器组变换矩阵
melcepst      - Mel cepstrum frontend for recogniser梅尔倒频谱前端识别
cep2pow       - Convert mel cepstram means & variances to power domain利用梅尔倒频谱均值和方差转换到功率域
pow2cep       - Convert power domain means & variances to mel cepstrum利用功率域转换到梅尔倒频谱均值和方差
ldatrace      - constrained Linear Discriminant Analysis to maximize trace(W\B)约束线性分析到最大限度跟踪

  

 

  
1
  
2
  
3
  
4
  
5
 

12 Signal Processing信号处理

ditherq       - Add dither and quantize a signal信号加抖动和量化（颤音？我自己猜想的）
filterbank    - Apply a bank of IIR filters to a signal对信号应用IIR过滤器
maxfilt       - Running maximum filter运行的最大值过滤器
meansqtf      - Output power of a filter with white noise input带有白噪声输入的波滤器的的功率输出
momfilt       - Generate running moments生成运行时刻
schmitt       - Pass a signal through a schmitt trigger信号通过施密特触发器
sigalign      - Align a clean refeence with a noisy signal对齐一个带有噪声信号的干净refeence
teager        - Calculate the Teager energy waveform Teager能量波形计算
v_addnoise    - Add noise to a signal at a chosen SNR 给信号加一个选择好的信噪比的噪声
v_findpeaks   - Find peaks in a signal or spectrum在一个信号或谱中找到峰
v_resample    - Resamples a signal: identical to MATLAB resample but removes filter transients重采样信号 和matlab自带重采样相同，但消除滤波器瞬变
v_windinfo    - Calculate window properties and figures of merit窗口性能和数字优点计算
v_windows     - Window function generation窗函数生成
zerocros      - Find interpolated zero crossings查找插值零点（零点）用buffer分片以后的波形数据可以作为输入参数，返回是波形数据的y=0时线性求的x点集合。（点处斜率正zerocros(y,'p') 负 zerocros(y,'n')  默认全部或者'b'）

  

 

  
1
  
2
  
3
  
4
  
5
  
6
  
7
  
8
  
9
  
10
  
11
  
12
  
13
  
14
 

13 Information Theory信息理论

huffman       - Generate Huffman code 生成哈夫曼编码
entropy       - Calculate entropy and conditional entropy熵和条件熵的计算

  

 

  
1
  
2
 

二、备注

版本：2014a

文章来源: qq912100926.blog.csdn.net，作者：海神之光，版权归原作者所有，如需转载，请联系作者。

原文链接：qq912100926.blog.csdn.net/article/details/112131366