📄 runkfda.m
字号:
% test case 2.2: n=40, minLocalSampleRatio = 0.25;
0.9079 0.8903 0.5833 0.9145 0.8375 0.7813
0.9558 0.9402 0.5763 0.9558 0.8917 0.8500
0.9333 0.8667 0.5125 0.9333 0.7875 0.8250
0.9873 0.9125 0.6410 0.9873 0.9125 0.8875
0.9200 0.9091 0.5395 0.9200 0.8625 0.8625
0.9324 0.8831 0.5897 0.9595 0.8000 0.8375
% test case 2.3: n=40, minPoseCnt = 3; minLocalSampleRatio = 0.20; defaultNNErr = 0.005;
% recRate recRt1 recRt2 recRt3 eig1 eig3
0.9079 0.8625 0.5750 0.8688 0.8375 0.7813
0.9558 0.9167 0.5667 0.9000 0.8917 0.8500
0.9333 0.8125 0.5125 0.8750 0.7875 0.8250
0.9873 0.9125 0.6250 0.9750 0.9125 0.8875
0.9200 0.8750 0.5125 0.8625 0.8625 0.8625
0.9324 0.8500 0.5750 0.8875 0.8000 0.8375
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% n=40;
wt1=ones(16);R1=640; wt2=ones(12);R2=600; wt3=ones(20);R3=1000;
recRate=[]; ct=[];
ids=[1:40];
for t=1:6
if t==1
poses = [1 2 4 5 9 10]; xPoses = [3 6 7 8];
elseif t==2
poses = [1 3 5 6 7 8 10]; xPoses = [2 4 9];
elseif t==3
poses = [1 2 3 4 6 7 8 9]; xPoses = [5 10];
elseif t==4
poses = [2 3 4 5 6 7 9 10]; xPoses = [1 8];
elseif t==5
poses = [1 3 4 5 6 8 9 10]; xPoses = [2 7];
else
poses = [1 2 4 5 7 8 9 10]; xPoses = [3 6];
end
[recRate(t,:), ct(t)]=testMultiModelFaceRecognition(Fm1, Fm2, Fm3, wt1, wt2, wt3, R1, R2, R3, ids, poses, xPoses, 'n');
%[recRate1, ct1]=testMultiModelFaceRecognition(Fm1, Fm2, Fm3, wt1, wt2, wt3, R1, R2, R3, ids, poses, xPoses, 'y');
end
recRate
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% test set (3)
% n=40, m=6 (known face per person)
% parameters:
% wt1=ones(16);R1=640; wt2=ones(12);R2=600; wt3=ones(20);R3=1000;
% minPoseCnt = 3; minLocalSampleRatio = 0.20; defaultNNErr = 0.005;
% test case 2.3: n=40, minPoseCnt = 3; minLocalSampleRatio = 0.20; defaultNNErr = 0.005;
% recRate recRt1 recRt2 recRt3 eig1 eig3 time
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
wt1=ones(16);R1=640; wt2=ones(12);R2=600; wt3=ones(20);R3=1000;
recRate=[]; ct=[]; m=6;
for k=1:80
x = randperm(10);
x1 = sort(x(1:m));
x2 = sort(x(m+1:10));
p(k,1:10)=[x1, x2];
end
offs=randperm(80);
testPoses = p(offs(21:30), 1:10)
testPoses=[
1 4 7 8 9 10 2 3 5 6
1 2 4 8 9 10 3 5 6 7
2 6 7 8 9 10 1 3 4 5
1 2 5 7 9 10 3 4 6 8
1 2 3 7 8 10 4 5 6 9
1 2 4 5 8 10 3 6 7 9
1 2 3 4 5 8 6 7 9 10
1 2 3 7 8 10 4 5 6 9
1 3 5 6 7 10 2 4 8 9
1 5 6 7 9 10 2 3 4 8
];
ids=[1:40]; m=6;
for t=1:10
poses = testPoses(t,1:m); xPoses=testPoses(t,m+1:10);
[recRate(t,:), ct(t)]=testMultiModelFaceRecognition(Fm1, Fm2, Fm3, wt1, wt2, wt3, R1, R2, R3, ids, poses, xPoses, 'n');
end
result=[recRate, ct']
r=results;
fprintf('\n CombRate R1 R2 R3 EigRate RecTime(sec)');
r(11,:)=mean(r); N=40*4;
for k=1:11
fprintf('\n %1.4f %1.4f %1.4f %1.4f %1.4f %1.4f ', r(k,1),r(k,2),r(k,3),r(k,4),r(k,6),r(k,7)/N);
end
% test case 2.3: n=40, minPoseCnt = 3; minLocalSampleRatio = 0.20; defaultNNErr = 0.005;
% recRate recRt1 recRt2 recRt3 eig1 eig3 time
results=[
0.9456 0.8438 0.5250 0.8750 0.8125 0.8375 37.4060
0.9392 0.8000 0.5250 0.8688 0.7750 0.8187 37.8130
0.9007 0.8688 0.4750 0.8500 0.8063 0.7875 37.4720
0.9145 0.8562 0.5313 0.8750 0.8375 0.8000 37.7070
0.8919 0.8125 0.5188 0.8250 0.7562 0.7500 36.7990
0.9067 0.8187 0.5313 0.8562 0.7562 0.8125 37.0970
0.9139 0.7937 0.5000 0.8625 0.7500 0.7500 36.9410
0.8919 0.8125 0.5188 0.8250 0.7562 0.7500 36.6440
0.9608 0.9063 0.5500 0.9125 0.8500 0.8625 38.0190
0.9346 0.9063 0.5500 0.8938 0.8250 0.8063 37.7850
];
===========================================================================
CombRate R1 R2 R3 EigRate RecTime(sec)
===========================================================================
0.9456 0.8438 0.5250 0.8750 0.8375 0.2338
0.9392 0.8000 0.5250 0.8688 0.8187 0.2363
0.9007 0.8688 0.4750 0.8500 0.7875 0.2342
0.9145 0.8562 0.5313 0.8750 0.8000 0.2357
0.8919 0.8125 0.5188 0.8250 0.7500 0.2300
0.9067 0.8187 0.5313 0.8562 0.8125 0.2319
0.9139 0.7937 0.5000 0.8625 0.7500 0.2309
0.8919 0.8125 0.5188 0.8250 0.7500 0.2290
0.9608 0.9063 0.5500 0.9125 0.8625 0.2376
0.9346 0.9063 0.5500 0.8938 0.8063 0.2362
---------------------------------------------------------------------------
0.9200 0.8419 0.5225 0.8644 0.7975 0.2336 (mean)
===========================================================================
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% test set (4)
% n=30, m=6 (known face per person)
% parameters:
% wt1=ones(16);R1=640; wt2=ones(12);R2=600; wt3=ones(20);R3=1000;
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
wt1=ones(16);R1=420; wt2=ones(12);R2=400; wt3=ones(20);R3=860;
recRate=[]; ct=[];
n=30; m=8;
% gen ids
testIds=randperm(40);
ids=sort(testIds(1:n));
% gen poses
for k=1:80
x = randperm(10);
x1 = sort(x(1:m));
x2 = sort(x(m+1:10));
p(k,1:10)=[x1, x2];
end
offs=randperm(80);
testPoses = p(offs(21:30), 1:10)
% test data:
ids=[1 2 3 4 5 7 8 9 10 11 12 13 14 15 16 17 19 20 21 22 23 24 29 31 34 36 37 38 39 40 ];
ids=[11:40];
testPoses =[
2 3 5 6 7 8 9 10 1 4
1 3 4 5 7 8 9 10 2 6
2 3 4 5 6 7 9 10 1 8
1 2 4 5 6 7 8 9 3 10
1 2 4 6 7 8 9 10 3 5
1 2 3 6 7 8 9 10 4 5
1 2 4 5 6 7 8 9 3 10
1 2 5 6 7 8 9 10 3 4
1 3 4 5 7 8 9 10 2 6
1 2 3 5 6 8 9 10 4 7
];
recRate=[];ct=[];
for t=1:10
poses = testPoses(t,1:m); xPoses=testPoses(t,m+1:10);
[recRate(t,:), ct(t)]=testMultiModelFaceRecognition(Fm1, Fm2, Fm3, wt1, wt2, wt3, R1, R2, R3, ids, poses, xPoses, 'n');
end
result=[recRate, ct']
% minPoseCnt = 3; minLocalSampleRatio = 0.30; defaultNNErr = 0.005;
% recRate recRt1 recRt2 recRt3 eig1 eig3 time
results=[
0.9310 0.8167 0.6667 0.9333 0.8333 0.9333 7.9520
0.9286 0.8500 0.7000 0.9000 0.8500 0.8500 7.8900
0.9492 0.9000 0.6667 0.9500 0.9333 0.9167 7.0150
0.9828 0.8833 0.5833 0.9500 0.8500 0.9333 7.5150
0.9298 0.8500 0.5333 0.9000 0.8167 0.8833 7.6080
0.8947 0.7833 0.5833 0.9000 0.7667 0.8333 7.1400
0.9828 0.8833 0.5833 0.9500 0.8500 0.9333 7.3120
0.8966 0.8167 0.5500 0.9000 0.8000 0.8667 7.3430
0.9286 0.8500 0.7000 0.9000 0.8500 0.8500 7.3110
0.9310 0.7833 0.6167 0.9333 0.7833 0.8333 7.4220
];
% minPoseCnt = 3; minLocalSampleRatio = 0.20; defaultNNErr = 0.005;
% recRate recRt1 recRt2 recRt3 eig1 eig3 time
r=[
0.9310 0.8333 0.6833 0.9333 0.8333 0.9333 7.1250
0.9643 0.8833 0.6833 0.9000 0.8500 0.8500 6.5770
0.9661 0.8833 0.6667 0.9667 0.9333 0.9167 7.1870
0.9828 0.8500 0.5833 0.9500 0.8500 0.9333 6.7490
0.9298 0.8333 0.5167 0.9000 0.8167 0.8833 6.0770
0.8772 0.7833 0.6000 0.9000 0.7667 0.8333 6.5150
0.9828 0.8500 0.5833 0.9500 0.8500 0.9333 6.0780
0.8966 0.8167 0.5500 0.9000 0.8000 0.8667 7.0300
0.9643 0.8833 0.6833 0.9000 0.8500 0.8500 5.8750
0.9310 0.8167 0.5833 0.9167 0.7833 0.8333 6.5770
];
===========================================================================
CombRate R1 R2 R3 EigRate RecTime(sec)
===========================================================================
0.9310 0.8333 0.6833 0.9333 0.9333 0.1187
0.9643 0.8833 0.6833 0.9000 0.8500 0.1096
0.9661 0.8833 0.6667 0.9667 0.9167 0.1198
0.9828 0.8500 0.5833 0.9500 0.9333 0.1125
0.9298 0.8333 0.5167 0.9000 0.8833 0.1013
0.8772 0.7833 0.6000 0.9000 0.8333 0.1086
0.9828 0.8500 0.5833 0.9500 0.9333 0.1013
0.8966 0.8167 0.5500 0.9000 0.8667 0.1172
0.9643 0.8833 0.6833 0.9000 0.8500 0.0979
0.9310 0.8167 0.5833 0.9167 0.8333 0.1096
---------------------------------------------------------------------------
0.9426 0.8433 0.6133 0.9217 0.8833 0.1097 (mean)
===========================================================================
total rec attempt 600
fprintf('\n CombRate R1 R2 R3 EigRate RecTime(sec)');
r(11,:)=mean(r); N=30*2;
for k=1:11
fprintf('\n %1.4f %1.4f %1.4f %1.4f %1.4f %1.4f ', r(k,1),r(k,2),r(k,3),r(k,4),r(k,6),r(k,7)/N);
end
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% test set (5)
% n=40, m=8 (known face per person)
% parameters:
% wt1=ones(16);R1=640; wt2=ones(12);R2=600; wt3=ones(20);R3=1000;
% minPoseCnt = 3; minLocalSampleRatio = 0.20; defaultNNErr = 0.005;
% test case 2.3: n=40, minPoseCnt = 3; minLocalSampleRatio = 0.20; defaultNNErr = 0.005;
% recRate recRt1 recRt2 recRt3 eig1 eig3 time
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
wt1=ones(16);R1=640; wt2=ones(12);R2=600; wt3=ones(20);R3=1000;
recRate=[]; ct=[]; m=8;
for k=1:80
x = randperm(10);
x1 = sort(x(1:m));
x2 = sort(x(m+1:10));
p(k,1:10)=[x1, x2];
end
offs=randperm(80);
testPoses = p(offs(21:30), 1:10)
testPoses=[
2 3 4 5 7 8 9 10 1 6
1 2 3 4 6 7 9 10 5 8
1 2 3 5 6 7 8 9 4 10
1 2 3 5 6 7 9 10 4 8
1 2 3 4 5 7 8 10 6 9
1 4 5 6 7 8 9 10 2 3
1 2 3 4 5 6 9 10 7 8
1 2 3 4 5 6 7 9 8 10
1 2 4 5 6 7 8 10 3 9
1 2 3 4 5 6 8 9 7 10
];
ids=[1:40]; m=8;
for t=1:10
poses = testPoses(t,1:m); xPoses=testPoses(t,m+1:10);
[recRate(t,:), ct(t)]=testMultiModelFaceRecognition(Fm1, Fm2, Fm3, wt1, wt2, wt3, R1, R2, R3, ids, poses, xPoses, 'n');
end
result=[recRate, ct']
result =[
0.9333 0.8375 0.4875 0.8750 0.8125 0.8000 0.2544
0.9605 0.8875 0.6250 0.9125 0.8750 0.8500 0.2597
0.9221 0.8625 0.4875 0.8875 0.7875 0.7750 0.2628
0.9744 0.9125 0.5625 0.9500 0.8875 0.8625 0.2310
0.8919 0.8000 0.5125 0.8250 0.7875 0.8000 0.2327
0.9733 0.9375 0.5500 0.9125 0.8625 0.8375 0.2195
0.9615 0.8625 0.5125 0.9375 0.8625 0.8250 0.2351
0.9620 0.8750 0.5750 0.9500 0.8500 0.8250 0.2293
0.9342 0.8875 0.5500 0.8875 0.9000 0.8875 0.2697
0.9221 0.8125 0.5125 0.8875 0.7875 0.8000 0.2484
];
r=result;
fprintf('\n CombRate R1 R2 R3 EigRate RecTime(sec)');
r(11,:)=mean(r);
for k=1:11
fprintf('\n %1.4f %1.4f %1.4f %1.4f %1.4f %1.4f ', r(k,1),r(k,2),r(k,3),r(k,4),r(k,6),r(k,7));
end
CombRate R1 R2 R3 EigRate RecTime(sec)
0.9333 0.8375 0.4875 0.8750 0.8000 0.2544
0.9605 0.8875 0.6250 0.9125 0.8500 0.2597
0.9221 0.8625 0.4875 0.8875 0.7750 0.2628
0.9744 0.9125 0.5625 0.9500 0.8625 0.2310
0.8919 0.8000 0.5125 0.8250 0.8000 0.2327
0.9733 0.9375 0.5500 0.9125 0.8375 0.2195
0.9615 0.8625 0.5125 0.9375 0.8250 0.2351
0.9620 0.8750 0.5750 0.9500 0.8250 0.2293
0.9342 0.8875 0.5500 0.8875 0.8875 0.2697
0.9221 0.8125 0.5125 0.8875 0.8000 0.2484
0.9435 0.8675 0.5375 0.9025 0.8263 0.2443 ⌨️ 快捷键说明
复制代码
Ctrl + C
搜索代码
Ctrl + F
全屏模式
F11
切换主题
Ctrl + Shift + D
显示快捷键
?
增大字号
Ctrl + =
减小字号
Ctrl + -