Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Register Log in

how to implement face detection using svm

Status
Not open for further replies.
E

emerson_11

Member level 2
Joined
Jan 23, 2016
Messages
44
Helped
1
Reputation
2
Reaction score
1
Trophy points
8
Activity points
355
i have taken features by HOG for around 10 sample images for both male and female and i have stored it in a matrix meas. Below is my code

Code: 
meas =    [a1;a2;a3;a4;a6;a7; a5;a8;a9;a10 ];
male = meas(1:5,:);
female  = meas(6:10,:);   
species = {'male';'male';'male';'male';'male';'female';'female';
 'female';'female';'female'};
save eme.mat meas

i have used meas matrix in svm classifier code as follows

data = vertcat(meas(1:5,:),meas(6:10,:));
c = [0;0;0;0;0;0;0;0;0;0];
groups = ismember(data,c);
 testFeat=[0.737383540372671,22,7180,7598,135.888399114184,0,0,0,0,0,0,0,0,0,0]
%   testFeat= measure(20,1);
 svmStruct = svmtrain(data,groups,'showplot',true);
classes = svmclassify(svmStruct,testFeat,'showplot',true);
   if classes==1
       msgbox('The given value belongs to male ');
   else
       msgbox('The given value belongs to female ');

  end

testfeat contains test image sample.am receiving the error as follows

Error in cr7 (line 451)
svmStruct = svmtrain(data,groups,'showplot',true);
Click to expand...

can anyone please correct my mistake?
 

Am doing a project in SVM based real time gender classification using fpga.
could anyone tell me what is nonlinear kernel?
 

Hi,

Can you do an internet search on "nonlinear kernel"?

It gives a lot of informations.
Even Wikipedia with "kernel method"..

I'm sure you did your research. So what information do you need?
We can't read all the documents and write it here.

Klaus
 

Since am doing real time classification using svm,am adviced to use non linear kernals for classification.
is dot product,polynomial,RBF is all non-linear kernel functions?
 

emerson_11 said:
Since am doing real time classification using svm,am adviced to use non linear kernals for classification.
is dot product,polynomial,RBF is all non-linear kernel functions?
Click to expand...

The question is not clear at all, but generally speaking: You can select many kernel functions, from linear to non linear, but each one has drawbacks, either in time required to make the regression ( namely, to find the parameters of kernel function itself ), or losing accuracy in results. The features of the image set that you have is what will dictate what to do. There is no way to discover that without making experiments.
 
Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top