classify2.m

function [out_maps,time_taken] = classify2( model, image, feat, Xt, Yt, Xv, Yv)
    numOfFeat = feat.length();
    perfL = zeros(1,numOfFeat);
    mdlL = cell(1,numOfFeat);

    tic();
    for i = 1:numOfFeat
        fimg = feat.get(i);
	[im_rows, im_cols, im_bands] = size( fimg );
	spectra = reshape( fimg, [im_rows*im_cols,im_bands] );
     	Xtrain = spectra( sub2ind([im_rows,im_cols],Xt(:,1),Xt(:,2)) ,:);
     	Xvalid = spectra( sub2ind([im_rows,im_cols],Xv(:,1),Xv(:,2)) ,:);
    	mdl = model();		
	mdl.train( Xtrain, Yt );
	[pred,~] = mdl.predict(Xvalid);
	perfL(i) = sum(pred' == Yv);
	mdlL{i} = mdl;
	%disp( sum(pred' == Yv) / length(Yv));
		
    end 
    [~,idx] = max(perfL);
    mdl = mdlL{idx};
    [out_image1,prob_map] = mdl.predictImg( feat.get(idx) );

    time1 = toc();
    tic();
    out_image2 = MRF( prob_map, Xv, Yv );
    time2 = toc();
    out_maps = {out_image1,out_image2};
    time_taken = [time1,time1+time2];
    
    numSuperPixs = [50,100,200,400,800,1600,3200,6400];

    for i = 1:length(numSuperPixs)
        tic();
        out_image3 = Super_MRF(prob_map,image,numSuperPixs(i),Xv,Yv);
        time3 = toc();  
        out_maps = [out_maps,out_image3];
	time_taken = [time_taken,time1+time3];
    end
	

end