Repository navigation
Expand file tree
/
Copy pathplot_results_graph.m
More file actions
221 lines (197 loc) · 7.62 KB
/
Copy pathplot_results_graph.m
File metadata and controls
221 lines (197 loc) · 7.62 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
clear
close all
%% SVM 10 classes
mean_rigid_SVM10 = [89.21 84.47 86.00 87.50];
mean_semi_SVM10 = [94.33 92.69 96.10 95.40];
mean_flex_SVM10 = [92.33 88.58 95.00 91.42];
std_rigid_SVM10 = [2.64 3.01 2.42 3.06];
std_semi_SVM10 = [2.98 2.52 1.22 1.07];
std_flex_SVM10 = [2.10 2.77 2.13 2.03];
training_time_20_SVM10 = (0.0367+0.0373+0.038)/3;
training_time_30_SVM10 = (0.0456+0.0452+0.0458)/3;
training_time_40_SVM10 = (0.0524+0.0486+0.0632)/3;
training_time_50_SVM10 = (0.0553+0.0526+0.0549)/3;
%% DCNN 10 classes
mean_rigid_DCNN10 = [92.25 88.69 90.73 92.50];
mean_semi_DCNN10 = [95.67 94.67 96.81 96.60];
mean_flex_DCNN10 = [94.12 90.33 92.44 93.38];
std_rigid_DCNN10 = [3.25 3.08 2.75 2.06];
std_semi_DCNN10 = [4.62 2.35 1.47 1.68];
std_flex_DCNN10 = [2.01 2.57 2.65 3.17];
training_time_20_DCNN10 = (4.3953+4.4472+4.3747)/3;
training_time_30_DCNN10 = (6.6187+6.5173+6.5117)/3;
training_time_40_DCNN10 = (8.9923+8.7988+8.8443)/3;
training_time_50_DCNN10 = (10.6040+10.5173+10.4892)/3;
%% SVM 15 classes
mean_rigid_SVM15 = [83 78.43 82.36 85.66];
mean_semi_SVM15 = [93.17 91.54 92.36 95.13];
mean_flex_SVM15 = [88.03 82.80 85.78 88.22];
std_rigid_SVM15 = [2.79 3.18 2.45 1.78];
std_semi_SVM15 = [1.56 1.76 1.48 1.71];
std_flex_SVM15 = [3.29 2.43 1.95 1.54];
training_time_20_SVM15 = (0.0694+0.0683+0.0651)/3;
training_time_30_SVM15 = (0.0706+0.0719+0.0723)/3;
training_time_40_SVM15 = (0.0889+0.0811+0.0853)/3;
training_time_50_SVM15 = (0.0990+0.0918+0.0992)/3;
%% DCNN 15 classes
mean_rigid_DCNN15 = [87.33 86.13 87.63 90.62];
mean_semi_DCNN15 = [95.17 93.87 95.78 95.73];
mean_flex_DCNN15 = [92.67 88.11 91.21 92.13];
std_rigid_DCNN15 = [3.46 3.23 3.47 2.5];
std_semi_DCNN15 = [2.38 1.88 1.58 1.33];
std_flex_DCNN15 = [2.76 3.38 2.29 2.12];
training_time_20_DCNN15 = (6.3659+6.3583+6.3746)/3;
training_time_30_DCNN15 = (9.5006+9.5411+9.6781)/3;
training_time_40_DCNN15 = (12.7467+12.7882+12.5969)/3;
training_time_50_DCNN15 = (15.6506+16.2281+15.9006)/3;
%% Ploting the results
% Recognition-Rate DCNN-SVM 10 classes
figure
subplot(2,2,1)
fig1 = barwitherr([std_rigid_SVM10; std_semi_SVM10; std_flex_SVM10],...
[mean_rigid_SVM10; mean_semi_SVM10; mean_flex_SVM10], 1);
legend('20 images','30 images','40 images','50 images')
set(gca,'XTickLabel',{'Rigid', 'Semi-rigid', 'Flexible', ' '})
ylabel('Mean Recognition Rate [%]')
ylim([50 100])
xlim([0.3 4.8])
title('DCNN-SVM (10 classes)')
color = summer(4);
set(fig1(1),'facecolor',color(1,:));
set(fig1(2),'facecolor',color(2,:));
set(fig1(3),'facecolor',color(3,:));
set(fig1(4),'facecolor',color(4,:));
grid on
% Recognition-Rate DCNN-SVM 15 classes
subplot(2,2,2)
fig2 = barwitherr([std_rigid_SVM15; std_semi_SVM15; std_flex_SVM15],...
[mean_rigid_SVM15; mean_semi_SVM15; mean_flex_SVM15], 1);
legend('20 images','30 images','40 images','50 images')
set(gca,'XTickLabel',{'Rigid', 'Semi-rigid', 'Flexible', ' '})
ylabel('Mean Recognition Rate [%]')
ylim([50 100])
xlim([0.3 4.8])
title('DCNN-SVM (15 classes)')
color = summer(4);
set(fig2(1),'facecolor',color(1,:));
set(fig2(2),'facecolor',color(2,:));
set(fig2(3),'facecolor',color(3,:));
set(fig2(4),'facecolor',color(4,:));
grid on
% Recognition-Rate DCNN 10 classes
subplot(2,2,3)
fig3 = barwitherr([std_rigid_DCNN10; std_semi_DCNN10; std_flex_DCNN10],...
[mean_rigid_DCNN10; mean_semi_DCNN10; mean_flex_DCNN10], 1);
legend('20 images','30 images','40 images','50 images')
set(gca,'XTickLabel',{'Rigid', 'Semi-rigid', 'Flexible', ' '})
ylabel('Mean Recognition Rate [%]')
ylim([50 100])
xlim([0.3 4.8])
title('DCNN (10 classes)')
color = summer(4);
set(fig3(1),'facecolor',color(1,:));
set(fig3(2),'facecolor',color(2,:));
set(fig3(3),'facecolor',color(3,:));
set(fig3(4),'facecolor',color(4,:));
grid on
% Recognition-Rate DCNN 15 classes
subplot(2,2,4)
fig4 = barwitherr([std_rigid_DCNN15; std_semi_DCNN15; std_flex_DCNN15],...
[mean_rigid_DCNN15; mean_semi_DCNN15; mean_flex_DCNN15], 1);
legend('20 images','30 images','40 images','50 images')
set(gca,'XTickLabel',{'Rigid', 'Semi-rigid', 'Flexible', ' '})
ylabel('Mean Recognition Rate [%]')
ylim([50 100])
xlim([0.3 4.8])
title('DCNN (15 classes)')
color = summer(4);
set(fig4(1),'facecolor',color(1,:));
set(fig4(2),'facecolor',color(2,:));
set(fig4(3),'facecolor',color(3,:));
set(fig4(4),'facecolor',color(4,:));
grid on
% Recognition-Rate global 10 classes rigid
figure
errorbar([20 30 40 50],mean_rigid_SVM10,std_rigid_SVM10,'b-','LineWidth', 2)
hold on
errorbar([20 30 40 50],mean_semi_SVM10,std_semi_SVM10,'r-','LineWidth', 2)
hold on
errorbar([20 30 40 50],mean_flex_SVM10,std_flex_SVM10,'g-','LineWidth', 2)
hold on
errorbar([20 30 40 50],mean_rigid_DCNN10,std_rigid_DCNN10,'b:','LineWidth', 2)
hold on
errorbar([20 30 40 50],mean_semi_DCNN10,std_semi_DCNN10,'r:','LineWidth', 2)
hold on
errorbar([20 30 40 50],mean_flex_DCNN10,std_flex_DCNN10,'g:','LineWidth', 2)
legend('Rigid (DCNN-SVM)','Semi-rigid (DCNN-SVM)','Flexible (DCNN-SVM)',...
'Rigid (DCNN)','Semi-rigid (DCNN)','Flexible (DCNN)', 'location','southeast')
ylabel('Mean Recognition Rate [%]')
xlabel('Number of images')
title('10 classes experiment')
ylim([50 105])
xlim([15 55])
grid on
% Recognition-Rate global 15 classes rigid
figure
errorbar([20 30 40 50],mean_rigid_SVM15,std_rigid_SVM15,'b-','LineWidth', 2)
hold on
errorbar([20 30 40 50],mean_semi_SVM15,std_semi_SVM15,'r-','LineWidth', 2)
hold on
errorbar([20 30 40 50],mean_flex_SVM15,std_flex_SVM15,'g-','LineWidth', 2)
hold on
errorbar([20 30 40 50],mean_rigid_DCNN15,std_rigid_DCNN15,'b:','LineWidth', 2)
hold on
errorbar([20 30 40 50],mean_semi_DCNN15,std_semi_DCNN15,'r:','LineWidth', 2)
hold on
errorbar([20 30 40 50],mean_flex_DCNN15,std_flex_DCNN15,'g:','LineWidth', 2)
legend('Rigid (DCNN-SVM)','Semi-rigid (DCNN-SVM)','Flexible (DCNN-SVM)',...
'Rigid (DCNN)','Semi-rigid (DCNN)','Flexible (DCNN)', 'location','southeast')
ylabel('Mean Recognition Rate [%]')
xlabel('Number of images')
title('15 classes experiment')
ylim([50 105])
xlim([15 55])
grid on
% Training time
figure
plot([20 30 40 50], [training_time_20_SVM10,training_time_30_SVM10,...
training_time_40_SVM10, training_time_50_SVM10],'o-', 'LineWidth', 2);
hold on
plot([20 30 40 50], [training_time_20_SVM15,training_time_30_SVM15,...
training_time_40_SVM15, training_time_50_SVM15],'^-','LineWidth', 2);
hold on
plot([20 30 40 50], [training_time_20_DCNN10,training_time_30_DCNN10,...
training_time_40_DCNN10, training_time_50_DCNN10],'s-', 'LineWidth', 2);
hold on
plot([20 30 40 50], [training_time_20_DCNN15,training_time_30_DCNN15,...
training_time_40_DCNN15, training_time_50_DCNN15],'*-', 'LineWidth', 2);
legend('DCNN-SVM (10 classes)','DCNN-SVM (15 classes)',...
'DCNN (10 classes)','DCNN (15 classes)', 'location','southeast')
ylabel('Training time [s]')
xlabel('Number of images')
ylim([0 16])
xlim([15 55])
grid on
%% two-levels experiment
mean_rigid_two = 98.78;
mean_semi_two = 97.94;
mean_flex_two = 95.92;
std_rigid_two = 1.15;
std_semi_two = 1.8;
std_flex_two = 1.67;
y = [mean_rigid_two, mean_semi_two, mean_flex_two]; %The data.
s = [std_rigid_two std_semi_two std_flex_two]; %The standard deviation.
fHand = figure;
aHand = axes('parent', fHand);
hold(aHand, 'on')
colors = summer(numel(y));
for i = 1:numel(y)
bar(i, y(i), 'parent', aHand, 'facecolor', colors(i,:));
end
set(gca, 'XTick', 1:numel(y), 'XTickLabel', {'Rigid', 'Semi-Rigid', 'Flexible'})
errorbar(y,s,'.k', 'LineWidth',1);
ylabel('Mean Recognition Rate [%]')
ylim([50 101])
xlim([0.3 3.7])
% title('First Level')
grid on