代码搜索：语音降噪
找到约 2,424 项符合「语音降噪」的源代码
代码结果 2,424
www.eeworm.com/read/206895/15286370
m mainfunction.m
function mainfunction()
wave=wavread('yinyue');
Xk0=wave(1:40000);
sound(Xk0,22050) % 播放原始语音信号
[XXh,Tx,DD,Td]=decomposition(Xk0,2);
Xk=reconstructed(XXh,Tx,DD,Td,2);
subplot(2,1,1),plot(Xk0),tit
www.eeworm.com/read/13871/284740
m inithmm.m
function hmm = inithmm(samples, M)
K = length(samples); %语音样本数
N = length(M); %状态数
hmm.N = N;
hmm.M = M;
% 初始概率矩阵
hmm.init = zeros(N,1);
hmm.init(1) = 1;
% 转移概率矩阵
hmm.trans=zeros(N
www.eeworm.com/read/13871/284742
m baum.m
function hmm = baum(hmm, samples)
mix = hmm.mix; %高斯混合
N = length(mix); %HMM状态数
K = length(samples); %语音样本数
SIZE = size(samples(1).data,2); %参数阶数
% 计算前向, 后向概率矩阵, 考虑多观察序列和下溢问题
disp(
www.eeworm.com/read/14686/402060
m inithmm.m
function hmm = inithmm(samples, M)
K = length(samples); %语音样本数
N = length(M); %状态数
hmm.N = N;
hmm.M = M;
% 初始概率矩阵
hmm.init = zeros(N,1);
hmm.init(1) = 1;
% 转移概率矩阵
hmm.trans=zeros(N
www.eeworm.com/read/14686/402062
m baum.m
function hmm = baum(hmm, samples)
mix = hmm.mix; %高斯混合
N = length(mix); %HMM状态数
K = length(samples); %语音样本数
SIZE = size(samples(1).data,2); %参数阶数
% 计算前向, 后向概率矩阵, 考虑多观察序列和下溢问题
disp(
www.eeworm.com/read/14761/407428
asv endpoint_watch.asv
function purewatch()
clear;
fclose('all');
frame_step=192;
temp=0;
% Fp_Speech=fopen('D:\test_kwspot\voice_converted\TargetFile\F_005.dat','r');
Fp_Speech=fopen('D:\PiPi_Demo\皮皮8k拆分后语音\8k_F_
www.eeworm.com/read/36061/897321
asm sacm_dvr1600.asm
//========================================================
// 文件名称： SACM_DVR1600.asm
// 功能描述： DVR1600语音播放用户接口程序
// 一般情况下仅需修改C_SYSTEMCLOCK和C_CLOCK_SET即可
// 维护记录： 2006-10-16 v1.0, by Qwerty
//==
www.eeworm.com/read/473523/6845410
m f7_5.m
%读入语音文件
fin=fopen('DR4_MLJH0_SX334.ADC','r');
x=fread(fin,'short');
fclose(fin);
%窗长
l=300;
step=100;
e_x=frame(x,'energy');
figure(1);
plot(e_x,'LineWidth',2);
xlabel('时间 t');
ylabel('帧输
www.eeworm.com/read/472368/6876114
m lsp.m
function [P_w,Q_w,H]=lsp(s,p,num)
%此函数的功能是求语音信号的线谱对参数
a=lpc_coefficients(s,p); %求线性预测系数
a=a'; %将a转化为行向量
p=length(a);
P_c=[1,-a-a(p:
www.eeworm.com/read/269960/11053312
asv inithmm.asv
function hmm = inithmm(samples, M)
% （ 改动 03－19)
%samples=wavread('101.wav');
%M=[3,3,3,3];
K = length(samples); %语音样本数
N = length(M); %状态数
hmm.N = N;
hmm.M = M;
% 初始概率矩阵
hmm.init = ze