need matlab code for features exctraction using MFCC

kareem dsp · Dec 20, 2012

hi all
i need matlab code for features exctraction using MFCC to use these featurs in dialect recognition

sarathhari · Dec 26, 2012

Hi i have a code and pdf for feature extraction using MFCC for speaker recognition. I will attach that please check it and use it if helpful.

Code:

AI=analoginput('winsound',1);
addchannel(AI,1);
set(AI,'samplerate',16000);
set(AI,'samplespertrigger',2*16000);
set(AI,'triggertype','immediate');
start(AI);
data=getdata(AI);
fs=16000;
%file = sprintf('%s%d.wav','database_new kar',i);
%%wavwrite(data,16000,file);

%% Computing MFCC Co-efficients..
    %% (1) Frame Blocking..
    N = 256;   % N point FFT
    M = 100;   % Overlapping

    NN = floor(N/2+1); %N/2
    nbFrames = ceil((length(data)-N)/M);
    Frames = zeros(nbFrames+1,N);
    for i = 0:nbFrames-1
        temp = data(i*M+1:i*M+N);
        Frames(i+1,1:N) = temp; 
    end

    % Last Frame..
    temp = zeros(1,N); 
    lastLength = length(data)- nbFrames*M;
    temp(1:lastLength) = data(nbFrames*M+1:(nbFrames*M +1 + lastLength-1));  
    Frames(nbFrames+1, 1:N) = temp;
    %% (2) Windowing..
    frameSize = size(Frames); 
    nbFrames = frameSize(1); 
    nbSamples = frameSize(2); 
 
    % Hamming window.. 
    w = hamming(nbSamples); 
    Windows = zeros(nbFrames,nbSamples);
    for i = 1:nbFrames
        temp = Frames(i,1:nbSamples); 
        Windows(i, 1:nbSamples) = w'.*temp; 
    end
    %% (3) Fourier Transform..
    ffts = fft(Windows');
    %% (4) Mel-frequency Wrapping..
    % (a) Calculate Power spectrum..
    PowSpecs = abs(ffts).^2;
    PowSpecs = PowSpecs(1:NN-1,:);
    % (b) Mel filter generation
    nof_c = 20; % Number of channels..
    df = fs/N;
    Nmax = N/2;
    fmax = fs/2;

    % Convert to mel scale..
    melmax = 2595*log10(1+fmax/700);

    melinc = melmax/(nof_c+1);

    melcenters = (1:nof_c).*melinc;

    % Convert to frequency scale.. 
    fcenters = 700*((10.^(melcenters./2595))-1);

    centerf = round(fcenters./df);

    startf = [1,centerf(1:nof_c-1)];
    stopf = [centerf(2:nof_c),Nmax];

    W = zeros(nof_c,Nmax);

    % Making filter..
    for i = 1:nof_c
        increment = 1.0/(centerf(i)-startf(i));
        for j = startf(i):centerf(i)
            W(i,j) = (j-startf(i))*increment;
        end
   
        decrement = 1.0/(stopf(i)-centerf(i));
        for j = centerf(i):stopf(i)
            W(i,j) = (j-centerf(i))*decrement;
        end 
    end
    % Normalising..
    for i = 1:nof_c
        W(i,:) = W(i,:)/sum(W(i,:));
    end
    
    % (c) Apply mel filters to Power spectrum coeffs..
    melPowSpecs = W*PowSpecs;
    % (d) MFCC calculations..
    melCeps = dct(log(melPowSpecs));
    melCeps(1,:) = [];
    
    load CodeBook
	dist_min = inf;
	spkr = 0;
	for ind = 1:length(codebook)
		single_cb =  codebook{ind};
		[M1, N1] = size(melCeps);
		[M2, N2] = size(single_cb);  
		dist_temp = zeros(N1,N2);
		if N1<N2
            repli = zeros(1,N2);
            for n1 = 1:N1
                dist_temp(n1,:) = sum((melCeps(:,n1+repli) - single_cb).^2,1);
            end
        else
            repli = zeros(1,N1);
            for n2 = 1:N2
                dist_temp(:,n2) = sum((melCeps - single_cb(:,n2+repli)).^2,1);
            end
        end
        dist_temp = sqrt(dist_temp);
		dist_val(ind) = sum(min(dist_temp,[],2))/size(dist_temp,1);
		if dist_val(ind) < dist_min
            dist_min = dist_val(ind);
			spkr = ind;
		end
	end
	msg = sprintf('The Speaker is found);
    	disp(msg);

pdf:

kareem dsp · Dec 26, 2012

sarathhari said:

Hi i have a code and pdf for feature extraction using MFCC for speaker recognition. I will attach that please check it and use it if helpful.

Code:

AI=analoginput('winsound',1);
addchannel(AI,1);
set(AI,'samplerate',16000);
set(AI,'samplespertrigger',2*16000);
set(AI,'triggertype','immediate');
start(AI);
data=getdata(AI);
fs=16000;
%file = sprintf('%s%d.wav','database_new kar',i);
%%wavwrite(data,16000,file);

%% Computing MFCC Co-efficients..
    %% (1) Frame Blocking..
    N = 256;   % N point FFT
    M = 100;   % Overlapping

    NN = floor(N/2+1); %N/2
    nbFrames = ceil((length(data)-N)/M);
    Frames = zeros(nbFrames+1,N);
    for i = 0:nbFrames-1
        temp = data(i*M+1:i*M+N);
        Frames(i+1,1:N) = temp; 
    end

    % Last Frame..
    temp = zeros(1,N); 
    lastLength = length(data)- nbFrames*M;
    temp(1:lastLength) = data(nbFrames*M+1:(nbFrames*M +1 + lastLength-1));  
    Frames(nbFrames+1, 1:N) = temp;
    %% (2) Windowing..
    frameSize = size(Frames); 
    nbFrames = frameSize(1); 
    nbSamples = frameSize(2); 
 
    % Hamming window.. 
    w = hamming(nbSamples); 
    Windows = zeros(nbFrames,nbSamples);
    for i = 1:nbFrames
        temp = Frames(i,1:nbSamples); 
        Windows(i, 1:nbSamples) = w'.*temp; 
    end
    %% (3) Fourier Transform..
    ffts = fft(Windows');
    %% (4) Mel-frequency Wrapping..
    % (a) Calculate Power spectrum..
    PowSpecs = abs(ffts).^2;
    PowSpecs = PowSpecs(1:NN-1,:);
    % (b) Mel filter generation
    nof_c = 20; % Number of channels..
    df = fs/N;
    Nmax = N/2;
    fmax = fs/2;

    % Convert to mel scale..
    melmax = 2595*log10(1+fmax/700);

    melinc = melmax/(nof_c+1);

    melcenters = (1:nof_c).*melinc;

    % Convert to frequency scale.. 
    fcenters = 700*((10.^(melcenters./2595))-1);

    centerf = round(fcenters./df);

    startf = [1,centerf(1:nof_c-1)];
    stopf = [centerf(2:nof_c),Nmax];

    W = zeros(nof_c,Nmax);

    % Making filter..
    for i = 1:nof_c
        increment = 1.0/(centerf(i)-startf(i));
        for j = startf(i):centerf(i)
            W(i,j) = (j-startf(i))*increment;
        end
   
        decrement = 1.0/(stopf(i)-centerf(i));
        for j = centerf(i):stopf(i)
            W(i,j) = (j-centerf(i))*decrement;
        end 
    end
    % Normalising..
    for i = 1:nof_c
        W(i,:) = W(i,:)/sum(W(i,:));
    end
    
    % (c) Apply mel filters to Power spectrum coeffs..
    melPowSpecs = W*PowSpecs;
    % (d) MFCC calculations..
    melCeps = dct(log(melPowSpecs));
    melCeps(1,:) = [];
    
    load CodeBook
	dist_min = inf;
	spkr = 0;
	for ind = 1:length(codebook)
		single_cb =  codebook{ind};
		[M1, N1] = size(melCeps);
		[M2, N2] = size(single_cb);  
		dist_temp = zeros(N1,N2);
		if N1<N2
            repli = zeros(1,N2);
            for n1 = 1:N1
                dist_temp(n1,:) = sum((melCeps(:,n1+repli) - single_cb).^2,1);
            end
        else
            repli = zeros(1,N1);
            for n2 = 1:N2
                dist_temp(:,n2) = sum((melCeps - single_cb(:,n2+repli)).^2,1);
            end
        end
        dist_temp = sqrt(dist_temp);
		dist_val(ind) = sum(min(dist_temp,[],2))/size(dist_temp,1);
		if dist_val(ind) < dist_min
            dist_min = dist_val(ind);
			spkr = ind;
		end
	end
	msg = sprintf('The Speaker is found);
    	disp(msg);

pdf:

thank you Sir for you interest but i can't see the PDF file

sarathhari · Dec 27, 2012

cant you see the link there....i am attaching it here plz check...in the code i have commented a portion showing MFCC Coefficients i think that will be helpful for you... i know it is a complex code just take the portion you wanted

pdfView attachment Speaker Recognition - Using MFCC.pdf

kareem dsp · Dec 27, 2012

thank you again sir

Ax's · Apr 25, 2013

sarathhari said:

Hi i have a code and pdf for feature extraction using MFCC for speaker recognition. I will attach that please check it and use it if helpful.

Code:

AI=analoginput('winsound',1);
addchannel(AI,1);
set(AI,'samplerate',16000);
set(AI,'samplespertrigger',2*16000);
set(AI,'triggertype','immediate');
start(AI);
data=getdata(AI);
fs=16000;
%file = sprintf('%s%d.wav','database_new kar',i);
%%wavwrite(data,16000,file);

%% Computing MFCC Co-efficients..
    %% (1) Frame Blocking..
    N = 256;   % N point FFT
    M = 100;   % Overlapping

    NN = floor(N/2+1); %N/2
    nbFrames = ceil((length(data)-N)/M);
    Frames = zeros(nbFrames+1,N);
    for i = 0:nbFrames-1
        temp = data(i*M+1:i*M+N);
        Frames(i+1,1:N) = temp; 
    end

    % Last Frame..
    temp = zeros(1,N); 
    lastLength = length(data)- nbFrames*M;
    temp(1:lastLength) = data(nbFrames*M+1:(nbFrames*M +1 + lastLength-1));  
    Frames(nbFrames+1, 1:N) = temp;
    %% (2) Windowing..
    frameSize = size(Frames); 
    nbFrames = frameSize(1); 
    nbSamples = frameSize(2); 
 
    % Hamming window.. 
    w = hamming(nbSamples); 
    Windows = zeros(nbFrames,nbSamples);
    for i = 1:nbFrames
        temp = Frames(i,1:nbSamples); 
        Windows(i, 1:nbSamples) = w'.*temp; 
    end
    %% (3) Fourier Transform..
    ffts = fft(Windows');
    %% (4) Mel-frequency Wrapping..
    % (a) Calculate Power spectrum..
    PowSpecs = abs(ffts).^2;
    PowSpecs = PowSpecs(1:NN-1,:);
    % (b) Mel filter generation
    nof_c = 20; % Number of channels..
    df = fs/N;
    Nmax = N/2;
    fmax = fs/2;

    % Convert to mel scale..
    melmax = 2595*log10(1+fmax/700);

    melinc = melmax/(nof_c+1);

    melcenters = (1:nof_c).*melinc;

    % Convert to frequency scale.. 
    fcenters = 700*((10.^(melcenters./2595))-1);

    centerf = round(fcenters./df);

    startf = [1,centerf(1:nof_c-1)];
    stopf = [centerf(2:nof_c),Nmax];

    W = zeros(nof_c,Nmax);

    % Making filter..
    for i = 1:nof_c
        increment = 1.0/(centerf(i)-startf(i));
        for j = startf(i):centerf(i)
            W(i,j) = (j-startf(i))*increment;
        end
   
        decrement = 1.0/(stopf(i)-centerf(i));
        for j = centerf(i):stopf(i)
            W(i,j) = (j-centerf(i))*decrement;
        end 
    end
    % Normalising..
    for i = 1:nof_c
        W(i,:) = W(i,:)/sum(W(i,:));
    end
    
    % (c) Apply mel filters to Power spectrum coeffs..
    melPowSpecs = W*PowSpecs;
    % (d) MFCC calculations..
    melCeps = dct(log(melPowSpecs));
    melCeps(1,:) = [];
    
    load CodeBook
	dist_min = inf;
	spkr = 0;
	for ind = 1:length(codebook)
		single_cb =  codebook{ind};
		[M1, N1] = size(melCeps);
		[M2, N2] = size(single_cb);  
		dist_temp = zeros(N1,N2);
		if N1<N2
            repli = zeros(1,N2);
            for n1 = 1:N1
                dist_temp(n1,:) = sum((melCeps(:,n1+repli) - single_cb).^2,1);
            end
        else
            repli = zeros(1,N1);
            for n2 = 1:N2
                dist_temp(:,n2) = sum((melCeps - single_cb(:,n2+repli)).^2,1);
            end
        end
        dist_temp = sqrt(dist_temp);
		dist_val(ind) = sum(min(dist_temp,[],2))/size(dist_temp,1);
		if dist_val(ind) < dist_min
            dist_min = dist_val(ind);
			spkr = ind;
		end
	end
	msg = sprintf('The Speaker is found);
    	disp(msg);

pdf:

this part is for what sir? please help me...:smile:
AI=analoginput('winsound',1);
addchannel(AI,1);
set(AI,'samplerate',16000);
set(AI,'samplespertrigger',2*16000);
set(AI,'triggertype','immediate');
start(AI);

sarathhari · Apr 26, 2013

Ax's said:
this part is for what sir? please help me...:smile:
AI=analoginput('winsound',1);
addchannel(AI,1);
set(AI,'samplerate',16000);
set(AI,'samplespertrigger',2*16000);
set(AI,'triggertype','immediate');
start(AI);

This section is to connect a recording device for example a mic into matlab environment. Analog input function selects the input device. add channel will create the device object. then triggering properties are varied using the set keyword.

89bibubibu · Apr 29, 2013

sarathhari said:

Hi i have a code and pdf for feature extraction using MFCC for speaker recognition. I will attach that please check it and use it if helpful.

Code:

AI=analoginput('winsound',1);
addchannel(AI,1);
set(AI,'samplerate',16000);
set(AI,'samplespertrigger',2*16000);
set(AI,'triggertype','immediate');
start(AI);
data=getdata(AI);
fs=16000;
%file = sprintf('%s%d.wav','database_new kar',i);
%%wavwrite(data,16000,file);

%% Computing MFCC Co-efficients..
    %% (1) Frame Blocking..
    N = 256;   % N point FFT
    M = 100;   % Overlapping

    NN = floor(N/2+1); %N/2
    nbFrames = ceil((length(data)-N)/M);
    Frames = zeros(nbFrames+1,N);
    for i = 0:nbFrames-1
        temp = data(i*M+1:i*M+N);
        Frames(i+1,1:N) = temp; 
    end

    % Last Frame..
    temp = zeros(1,N); 
    lastLength = length(data)- nbFrames*M;
    temp(1:lastLength) = data(nbFrames*M+1:(nbFrames*M +1 + lastLength-1));  
    Frames(nbFrames+1, 1:N) = temp;
    %% (2) Windowing..
    frameSize = size(Frames); 
    nbFrames = frameSize(1); 
    nbSamples = frameSize(2); 
 
    % Hamming window.. 
    w = hamming(nbSamples); 
    Windows = zeros(nbFrames,nbSamples);
    for i = 1:nbFrames
        temp = Frames(i,1:nbSamples); 
        Windows(i, 1:nbSamples) = w'.*temp; 
    end
    %% (3) Fourier Transform..
    ffts = fft(Windows');
    %% (4) Mel-frequency Wrapping..
    % (a) Calculate Power spectrum..
    PowSpecs = abs(ffts).^2;
    PowSpecs = PowSpecs(1:NN-1,:);
    % (b) Mel filter generation
    nof_c = 20; % Number of channels..
    df = fs/N;
    Nmax = N/2;
    fmax = fs/2;

    % Convert to mel scale..
    melmax = 2595*log10(1+fmax/700);

    melinc = melmax/(nof_c+1);

    melcenters = (1:nof_c).*melinc;

    % Convert to frequency scale.. 
    fcenters = 700*((10.^(melcenters./2595))-1);

    centerf = round(fcenters./df);

    startf = [1,centerf(1:nof_c-1)];
    stopf = [centerf(2:nof_c),Nmax];

    W = zeros(nof_c,Nmax);

    % Making filter..
    for i = 1:nof_c
        increment = 1.0/(centerf(i)-startf(i));
        for j = startf(i):centerf(i)
            W(i,j) = (j-startf(i))*increment;
        end
   
        decrement = 1.0/(stopf(i)-centerf(i));
        for j = centerf(i):stopf(i)
            W(i,j) = (j-centerf(i))*decrement;
        end 
    end
    % Normalising..
    for i = 1:nof_c
        W(i,:) = W(i,:)/sum(W(i,:));
    end
    
    % (c) Apply mel filters to Power spectrum coeffs..
    melPowSpecs = W*PowSpecs;
    % (d) MFCC calculations..
    melCeps = dct(log(melPowSpecs));
    melCeps(1,:) = [];
    
    load CodeBook
	dist_min = inf;
	spkr = 0;
	for ind = 1:length(codebook)
		single_cb =  codebook{ind};
		[M1, N1] = size(melCeps);
		[M2, N2] = size(single_cb);  
		dist_temp = zeros(N1,N2);
		if N1<N2
            repli = zeros(1,N2);
            for n1 = 1:N1
                dist_temp(n1,:) = sum((melCeps(:,n1+repli) - single_cb).^2,1);
            end
        else
            repli = zeros(1,N1);
            for n2 = 1:N2
                dist_temp(:,n2) = sum((melCeps - single_cb(:,n2+repli)).^2,1);
            end
        end
        dist_temp = sqrt(dist_temp);
		dist_val(ind) = sum(min(dist_temp,[],2))/size(dist_temp,1);
		if dist_val(ind) < dist_min
            dist_min = dist_val(ind);
			spkr = ind;
		end
	end
	msg = sprintf('The Speaker is found);
    	disp(msg);

pdf:

Hi sir,

i tried run your code but it seems there appear some "cut-off" triangular filters(normalized,image - MFCCfilters(Sir's Code) ) instead of the triangular filter ouput(only image) that i took from one internet source (normalized,bottom image of image - TriFilterBank(From other source)). Can sir explain to me about this ? Thanks. =)

sarathhari · Apr 29, 2013

89bibubibu said:
Hi sir,

i tried run your code but it seems there appear some "cut-off" triangular filters(normalized,image - MFCCfilters(Sir's Code) ) instead of the triangular filter ouput(only image) that i took from one internet source (normalized,bottom image of image - TriFilterBank(From other source)). Can sir explain to me about this ? Thanks. =)

Actually the code written above is for voice recognition and not for image recognition that may be the problem with you

89bibubibu · Apr 29, 2013

sarathhari said:
Actually the code written above is for voice recognition and not for image recognition that may be the problem with you

No sir, you misunderstood what i meant. The images that i mentioned and attached are the MFCC filter graph plotted. Not related to image recognition at all. I tried your code to plot out the triangular filters but it seems some "cut-off" triangular filters there...

der14 · Sep 11, 2013

sir, if this is the code for feature extraction using mfcc, please could you give me the code for feature extraction using fft (i have to input the wav signal using wavread)
please could you help me.

Parshu · Sep 27, 2013

sarathhari said:

Hi i have a code and pdf for feature extraction using MFCC for speaker recognition. I will attach that please check it and use it if helpful.

Code:

AI=analoginput('winsound',1);
addchannel(AI,1);
set(AI,'samplerate',16000);
set(AI,'samplespertrigger',2*16000);
set(AI,'triggertype','immediate');
start(AI);
data=getdata(AI);
fs=16000;
%file = sprintf('%s%d.wav','database_new kar',i);
%%wavwrite(data,16000,file);

%% Computing MFCC Co-efficients..
    %% (1) Frame Blocking..
    N = 256;   % N point FFT
    M = 100;   % Overlapping

    NN = floor(N/2+1); %N/2
    nbFrames = ceil((length(data)-N)/M);
    Frames = zeros(nbFrames+1,N);
    for i = 0:nbFrames-1
        temp = data(i*M+1:i*M+N);
        Frames(i+1,1:N) = temp; 
    end

    % Last Frame..
    temp = zeros(1,N); 
    lastLength = length(data)- nbFrames*M;
    temp(1:lastLength) = data(nbFrames*M+1:(nbFrames*M +1 + lastLength-1));  
    Frames(nbFrames+1, 1:N) = temp;
    %% (2) Windowing..
    frameSize = size(Frames); 
    nbFrames = frameSize(1); 
    nbSamples = frameSize(2); 
 
    % Hamming window.. 
    w = hamming(nbSamples); 
    Windows = zeros(nbFrames,nbSamples);
    for i = 1:nbFrames
        temp = Frames(i,1:nbSamples); 
        Windows(i, 1:nbSamples) = w'.*temp; 
    end
    %% (3) Fourier Transform..
    ffts = fft(Windows');
    %% (4) Mel-frequency Wrapping..
    % (a) Calculate Power spectrum..
    PowSpecs = abs(ffts).^2;
    PowSpecs = PowSpecs(1:NN-1,:);
    % (b) Mel filter generation
    nof_c = 20; % Number of channels..
    df = fs/N;
    Nmax = N/2;
    fmax = fs/2;

    % Convert to mel scale..
    melmax = 2595*log10(1+fmax/700);

    melinc = melmax/(nof_c+1);

    melcenters = (1:nof_c).*melinc;

    % Convert to frequency scale.. 
    fcenters = 700*((10.^(melcenters./2595))-1);

    centerf = round(fcenters./df);

    startf = [1,centerf(1:nof_c-1)];
    stopf = [centerf(2:nof_c),Nmax];

    W = zeros(nof_c,Nmax);

    % Making filter..
    for i = 1:nof_c
        increment = 1.0/(centerf(i)-startf(i));
        for j = startf(i):centerf(i)
            W(i,j) = (j-startf(i))*increment;
        end
   
        decrement = 1.0/(stopf(i)-centerf(i));
        for j = centerf(i):stopf(i)
            W(i,j) = (j-centerf(i))*decrement;
        end 
    end
    % Normalising..
    for i = 1:nof_c
        W(i,:) = W(i,:)/sum(W(i,:));
    end
    
    % (c) Apply mel filters to Power spectrum coeffs..
    melPowSpecs = W*PowSpecs;
    % (d) MFCC calculations..
    melCeps = dct(log(melPowSpecs));
    melCeps(1,:) = [];
    
    load CodeBook
	dist_min = inf;
	spkr = 0;
	for ind = 1:length(codebook)
		single_cb =  codebook{ind};
		[M1, N1] = size(melCeps);
		[M2, N2] = size(single_cb);  
		dist_temp = zeros(N1,N2);
		if N1<N2
            repli = zeros(1,N2);
            for n1 = 1:N1
                dist_temp(n1,:) = sum((melCeps(:,n1+repli) - single_cb).^2,1);
            end
        else
            repli = zeros(1,N1);
            for n2 = 1:N2
                dist_temp(:,n2) = sum((melCeps - single_cb(:,n2+repli)).^2,1);
            end
        end
        dist_temp = sqrt(dist_temp);
		dist_val(ind) = sum(min(dist_temp,[],2))/size(dist_temp,1);
		if dist_val(ind) < dist_min
            dist_min = dist_val(ind);
			spkr = ind;
		end
	end
	msg = sprintf('The Speaker is found);
    	disp(msg);

pdf:

Thank you sir for this code. I am also working of automatic speaker recognition with MFCC. Will you please share the complete code us for speaker recognition.??

aphnaseem · Sep 28, 2013

sarathhari said:
Actually the code written above is for voice recognition and not for image recognition that may be the problem with you

APHNASEEM:

sir what is cookbook, is it file or function can you give me plz,..

vicky2889 · Oct 20, 2013

hello sir i tried this code but can't understand ...... this codebook.......... and also this is not working
please tell me

kareem dsp · Oct 20, 2013

i am sorry sir the above code is not mine their is some mistake i uploaded another one which is in fact not mine i mean i didn't write i download it from internet site

gstekboy · Oct 27, 2013

Can i have lpc and plp code.

Syahroni Hidayat · Nov 15, 2013

is there any different between code of feature extraction using MFCC for speaker recognition and speech recognition?

- - - Updated - - -

sarathhari said:

Hi i have a code and pdf for feature extraction using MFCC for speaker recognition. I will attach that please check it and use it if helpful.

Code:

AI=analoginput('winsound',1);
addchannel(AI,1);
set(AI,'samplerate',16000);
set(AI,'samplespertrigger',2*16000);
set(AI,'triggertype','immediate');
start(AI);
data=getdata(AI);
fs=16000;
%file = sprintf('%s%d.wav','database_new kar',i);
%%wavwrite(data,16000,file);

%% Computing MFCC Co-efficients..
    %% (1) Frame Blocking..
    N = 256;   % N point FFT
    M = 100;   % Overlapping

    NN = floor(N/2+1); %N/2
    nbFrames = ceil((length(data)-N)/M);
    Frames = zeros(nbFrames+1,N);
    for i = 0:nbFrames-1
        temp = data(i*M+1:i*M+N);
        Frames(i+1,1:N) = temp; 
    end

    % Last Frame..
    temp = zeros(1,N); 
    lastLength = length(data)- nbFrames*M;
    temp(1:lastLength) = data(nbFrames*M+1:(nbFrames*M +1 + lastLength-1));  
    Frames(nbFrames+1, 1:N) = temp;
    %% (2) Windowing..
    frameSize = size(Frames); 
    nbFrames = frameSize(1); 
    nbSamples = frameSize(2); 
 
    % Hamming window.. 
    w = hamming(nbSamples); 
    Windows = zeros(nbFrames,nbSamples);
    for i = 1:nbFrames
        temp = Frames(i,1:nbSamples); 
        Windows(i, 1:nbSamples) = w'.*temp; 
    end
    %% (3) Fourier Transform..
    ffts = fft(Windows');
    %% (4) Mel-frequency Wrapping..
    % (a) Calculate Power spectrum..
    PowSpecs = abs(ffts).^2;
    PowSpecs = PowSpecs(1:NN-1,:);
    % (b) Mel filter generation
    nof_c = 20; % Number of channels..
    df = fs/N;
    Nmax = N/2;
    fmax = fs/2;

    % Convert to mel scale..
    melmax = 2595*log10(1+fmax/700);

    melinc = melmax/(nof_c+1);

    melcenters = (1:nof_c).*melinc;

    % Convert to frequency scale.. 
    fcenters = 700*((10.^(melcenters./2595))-1);

    centerf = round(fcenters./df);

    startf = [1,centerf(1:nof_c-1)];
    stopf = [centerf(2:nof_c),Nmax];

    W = zeros(nof_c,Nmax);

    % Making filter..
    for i = 1:nof_c
        increment = 1.0/(centerf(i)-startf(i));
        for j = startf(i):centerf(i)
            W(i,j) = (j-startf(i))*increment;
        end
   
        decrement = 1.0/(stopf(i)-centerf(i));
        for j = centerf(i):stopf(i)
            W(i,j) = (j-centerf(i))*decrement;
        end 
    end
    % Normalising..
    for i = 1:nof_c
        W(i,:) = W(i,:)/sum(W(i,:));
    end
    
    % (c) Apply mel filters to Power spectrum coeffs..
    melPowSpecs = W*PowSpecs;
    % (d) MFCC calculations..
    melCeps = dct(log(melPowSpecs));
    melCeps(1,:) = [];
    
    load CodeBook
	dist_min = inf;
	spkr = 0;
	for ind = 1:length(codebook)
		single_cb =  codebook{ind};
		[M1, N1] = size(melCeps);
		[M2, N2] = size(single_cb);  
		dist_temp = zeros(N1,N2);
		if N1<N2
            repli = zeros(1,N2);
            for n1 = 1:N1
                dist_temp(n1,:) = sum((melCeps(:,n1+repli) - single_cb).^2,1);
            end
        else
            repli = zeros(1,N1);
            for n2 = 1:N2
                dist_temp(:,n2) = sum((melCeps - single_cb(:,n2+repli)).^2,1);
            end
        end
        dist_temp = sqrt(dist_temp);
		dist_val(ind) = sum(min(dist_temp,[],2))/size(dist_temp,1);
		if dist_val(ind) < dist_min
            dist_min = dist_val(ind);
			spkr = ind;
		end
	end
	msg = sprintf('The Speaker is found);
    	disp(msg);

pdf:

is there any different between code of feature extraction using MFCC for speaker recognition and speech recognition?

Adem Ba · Mar 8, 2014

hi can u help me how you created codebook in speech recognition

dababseh · Mar 14, 2014

hello
can u help me to run this code,and making my coodbook?

parisah · Apr 4, 2014

hello sir!
i tried to execute your code but there is an error in loading the codebook!
can u please help me solvig the problem

need matlab code for features exctraction using MFCC

Junior Member level 1

Junior Member level 3

Attachments

Junior Member level 1

Junior Member level 3

Junior Member level 1

Newbie level 3

Junior Member level 3

Newbie level 2

Attachments

Junior Member level 3

Newbie level 2

Newbie level 4

Newbie level 1

Newbie level 1

Newbie level 1

Junior Member level 1

Member level 5

Newbie level 1

Newbie level 2

Newbie level 1

Newbie level 1

Similar threads

Privacy & Transparency

Privacy & Transparency