/
eigen_faces_pca.py
executable file
·169 lines (145 loc) · 4.9 KB
/
eigen_faces_pca.py
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
import numpy as np
import cv2 as cv
import os
import matplotlib.pyplot as plt
from numpy.linalg import svd
# task 1
def display_images(img, row, col, ftitle=""):
if row == 1 and col == 1:
plt.imshow(cv.resize(img, (100,100)))
plt.title(ftitle)
elif col == 2 :
figs, axes = plt.subplots(row,col)
figs.set_size_inches(50, 10)
figs.suptitle(ftitle)
for i in range(row):
for j in range(col):
axes[i, j].imshow(cv.resize(img[(i * 2 + j)], (100, 100)), cmap=plt.cm.gray)
else:
figs, axes = plt.subplots(row, col)
figs.set_size_inches(10, 5)
figs.suptitle(ftitle)
for i in range(row):
for j in range(col):
axes[i,j].imshow(cv.resize(img[(i*row + j)], (100,100)), cmap=plt.cm.gray)
plt.show()
train_images = "./Eigenfaces/Train/"
test_images = "./Eigenfaces/Test/"
images = []
train_img_names = []
for img_file in os.listdir(train_images):
file_name = os.path.join(train_images, img_file)
train_img_names.append(img_file.split("_")[0])
img = (cv.imread(file_name, 0)) / 255.
images.append(img)
images = np.array(images)
img_vec = images.reshape(images.shape[0],-1)
mean_face = np.mean(images, axis=0)
zerom_img = img_vec - mean_face.reshape(-1)
sd = np.std(zerom_img)
zerom_img /= sd
#display_images(images, 5, 5)
#display_images(mean_face, 1,1)
cov_mat = np.cov(zerom_img)
egval, egvec = np.linalg.eig(cov_mat)
egfaces = np.transpose(np.dot(np.transpose(zerom_img), egvec))
nm= np.linalg.norm(egfaces, axis=1)
nm= nm.reshape(-1,1)
egfaces /= nm
egfaces,s,v = svd(np.transpose(zerom_img), full_matrices=False)
eg_index = np.argsort(egval)[::-1]
#display_images(egfaces.reshape(images.shape),5,5)
# task 2
components = [2, 5, 15]
eg_faceset = []
prj_face = []
for comp in components:
top_egfaces = []
for k in range(comp):
top_egfaces.append(egfaces[eg_index[k]])
eg_faceset.append(top_egfaces)
top_egfaces = np.array(top_egfaces)
projected_faces = np.dot(zerom_img, np.transpose(top_egfaces))
prj_face.append(projected_faces)
rec_images = mean_face.reshape(-1) + np.dot(projected_faces, top_egfaces)
rec_images /= np.std(rec_images)
#display_images(rec_images.reshape(images.shape),5,5)
# task 3
def get_egfaces(components = [2, 5, 15]):
eg_faceset = []
prj_face = []
for comp in components:
top_egfaces = []
for k in range(comp):
top_egfaces.append(egfaces[eg_index[k]])
eg_faceset.append(top_egfaces)
top_egfaces = np.array(top_egfaces)
projected_faces = np.dot(zerom_img, np.transpose(top_egfaces))
prj_face.append(projected_faces)
return eg_faceset, prj_face
timages = []
tlabels = []
blnkimg = np.ones((images.shape[1],images.shape[2]))
#blnkimg /= 2.
for img_file in os.listdir(test_images):
file_name = os.path.join(test_images, img_file)
img = (cv.imread(file_name, 0)) / 255.
img = cv.resize(img,(images.shape[1],images.shape[2]))
timages.append(img)
if img_file.split("_")[0] in train_img_names:
tlabels.append(1)
else:
tlabels.append(-1)
tlabels = np.array(tlabels)
tlabels[1:]
tlabels[-7:] = -1
timages = np.array(timages)
zm_timg = timages - mean_face
zm_timg /= sd
timg_vec = zm_timg.reshape(timages.shape[0],-1)
c= [25]
eg_faceset, prj_face = get_egfaces(c)
dist = []
for fin, egf in enumerate(eg_faceset):
timg_proj = np.dot(timg_vec, np.transpose(egf))
pred = []
for index, tmp in enumerate(timg_proj):
t1 = prj_face[fin]
t2 = tmp
e_dist = np.sqrt(np.sum(np.square(prj_face[fin] - tmp), axis=1))
dist.append(np.min(e_dist))
if np.min(e_dist) < 12.5*c[fin]:
pred.append(np.argmin(e_dist).tolist())
else:
pred.append(-1)
pred_imgs = []
for index, p in enumerate(pred):
pred_imgs.append(timages[index])
if p != -1:
pred_imgs.append(images[p])
else:
#pred_imgs.append(-1)
pred_imgs.append(blnkimg)
pred_imgs = np.array(pred_imgs)
display_images(pred_imgs,timages.shape[0],2)
# task 4
all_com = np.arange(1,images.shape[0])
eg_faceset, prj_face = get_egfaces(all_com)
x_val = []
error = []
for fin, egf in enumerate(eg_faceset):
timg_proj = np.dot(timg_vec, np.transpose(egf))
pred = []
for index, tmp in enumerate(timg_proj):
e_dist = np.sqrt(np.sum(np.square(prj_face[fin] - tmp), axis=1))
if np.min(e_dist) < 13.5*all_com[fin]:
pred.append(1)
else:
pred.append(-1)
x_val.append(all_com[fin])
error.append((np.sum(tlabels != np.array(pred))/timages.shape[0])*100)
plt.plot(x_val, error)
plt.ylabel("Error rate")
plt.xlabel("k value")
plt.ylim(0,100)
plt.show()