-
Notifications
You must be signed in to change notification settings - Fork 0
/
RED_Dataset.py
262 lines (219 loc) · 9.04 KB
/
RED_Dataset.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
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
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
# coding: utf-8
import cv2
from torch.utils.data import Dataset
import Transform_Model as TM
import random
# import dlib
import numpy as np
from PIL import Image
import torchvision.transforms.functional as tf
from torchvision import transforms
import torch
class FaceDataset(Dataset):
def __init__(self, txt_path, transform = None):
fh= open(txt_path, 'r')
clean_imgs = []
adv_imgs = []
for line in fh:
line = line.rstrip()
words = line.split()
clean_imgs.append(words[0])
adv_imgs.append(words[1])
self.clean_imgs = clean_imgs # 最主要就是要生成这个list， 然后DataLoader中给index，通过getitem读取图片数据
self.adv_imgs = adv_imgs
self.transform = transform
def rotation(self, image1, image2):
# get a random angle range from (-180, 180)
angle = transforms.RandomRotation.get_params([-180, 180])
# same angle rotation for image1 and image2
image1 = image1.rotate(angle)
image2 = image2.rotate(angle)
image1 = tf.to_tensor(image1)
image2 = tf.to_tensor(image2)
return image1, image2
def flip(self, image1, image2):
# 50% prob to horizontal flip and vertical flip
if random.random() > 0.5:
image1 = tf.hflip(image1)
image2 = tf.hflip(image2)
if random.random() > 0.5:
image1 = tf.vflip(image1)
image2 = tf.vflip(image2)
image1 = tf.to_tensor(image1)
image2 = tf.to_tensor(image2)
return image1, image2
def __getitem__(self, index):
clean_address = self.clean_imgs[index]
adv_address = self.adv_imgs[index]
clean_img = TM.preprocess_image(cv2.imread(clean_address))
adv_img = TM.preprocess_image(cv2.imread(adv_address))
# if self.transform is not None:
# clean_img = self.transform(clean_img)
# adv_img = self.transform(adv_img)
if self.transform == 'rotation':
clean_img, adv_img = self.rotation(clean_img, adv_img)
elif self.transform == 'flip':
clean_img, adv_img = self.flip(clean_img, adv_img)
else:
clean_img = tf.to_tensor(clean_img)
adv_img = tf.to_tensor(adv_img)
return clean_img, adv_img
def __len__(self):
return len(self.clean_imgs)
class FaceDatasetTransformTest(Dataset):
def __init__(self, txt_path, transform = None):
fh= open(txt_path, 'r')
clean_imgs = []
adv_imgs = []
for line in fh:
line = line.rstrip()
words = line.split()
clean_imgs.append(words[0])
adv_imgs.append(words[1])
self.clean_imgs = clean_imgs # 最主要就是要生成这个list， 然后DataLoader中给index，通过getitem读取图片数据
self.adv_imgs = adv_imgs
self.transform = transform
def rotation(self, image1, image2):
# get a random angle range from (-180, 180)
angle = transforms.RandomRotation.get_params([-180, 180])
# same angle rotation for image1 and image2
image1 = image1.rotate(angle)
image2 = image2.rotate(angle)
image1 = tf.to_tensor(image1)
image2 = tf.to_tensor(image2)
return image1, image2
def flip(self, image1, image2):
# 50% prob to horizontal flip and vertical flip
if random.random() > 0.5:
image1 = tf.hflip(image1)
image2 = tf.hflip(image2)
if random.random() > 0.5:
image1 = tf.vflip(image1)
image2 = tf.vflip(image2)
image1 = tf.to_tensor(image1)
image2 = tf.to_tensor(image2)
return image1, image2
def hflip(self, image1, image2):
image1 = tf.hflip(image1)
image2 = tf.hflip(image2)
image1 = tf.to_tensor(image1)
image2 = tf.to_tensor(image2)
return image1, image2
def vflip(self, image1, image2):
image1 = tf.vflip(image1)
image2 = tf.vflip(image2)
image1 = tf.to_tensor(image1)
image2 = tf.to_tensor(image2)
return image1, image2
def rotation_new(self, image1, image2):
if random.random() > 0.5:
angle = transforms.RandomRotation.get_params([40, 50])
else:
angle = transforms.RandomRotation.get_params([-50, -40])
image1 = image1.rotate(angle)
image2 = image2.rotate(angle)
image1 = tf.to_tensor(image1)
image2 = tf.to_tensor(image2)
return image1, image2
def __getitem__(self, index):
clean_address = self.clean_imgs[index]
adv_address = self.adv_imgs[index]
clean_img = TM.preprocess_image(cv2.imread(clean_address))
adv_img = TM.preprocess_image(cv2.imread(adv_address))
# if self.transform is not None:
# clean_img = self.transform(clean_img)
# adv_img = self.transform(adv_img)
if self.transform == 'rotation':
clean_img_transform, adv_img_transform = self.rotation(clean_img, adv_img)
elif self.transform == 'flip':
clean_img_transform, adv_img_transform = self.flip(clean_img, adv_img)
elif self.transform == 'hflip':
clean_img_transform, adv_img_transform = self.hflip(clean_img, adv_img)
elif self.transform == 'vflip':
clean_img_transform, adv_img_transform = self.vflip(clean_img, adv_img)
elif self.transform == 'rotation_new':
clean_img_transform, adv_img_transform = self.rotation_new(clean_img, adv_img)
clean_img = tf.to_tensor(clean_img)
adv_img = tf.to_tensor(adv_img)
return clean_img, adv_img, clean_img_transform, adv_img_transform
def __len__(self):
return len(self.clean_imgs)
class Labeled_FaceDataset(Dataset):
def __init__(self, txt_path, label):
fh = open(txt_path, 'r')
clean_imgs = []
adv_imgs = []
# labels = []
for line in fh:
line = line.rstrip()
words = line.split()
clean_imgs.append(words[0])
adv_imgs.append(words[1])
# labels.append(label)
self.clean_imgs = clean_imgs # 最主要就是要生成这个list， 然后DataLoader中给index，通过getitem读取图片数据
self.adv_imgs = adv_imgs
self.label = label
def __getitem__(self, index):
clean_address = self.clean_imgs[index]
adv_address = self.adv_imgs[index]
clean_img = TM.preprocess_image(cv2.imread(clean_address))
adv_img = TM.preprocess_image(cv2.imread(adv_address))
# print(clean_img.type)
clean_img = tf.to_tensor(clean_img)
adv_img = tf.to_tensor(adv_img)
return torch.cat((adv_img-clean_img, clean_img),0), self.label
def __len__(self):
return len(self.clean_imgs)
class Labeled_FaceDataset_new(Dataset):
def __init__(self, txt_path, label):
fh = open(txt_path, 'r')
clean_imgs = []
adv_imgs = []
# labels = []
for line in fh:
line = line.rstrip()
words = line.split()
clean_imgs.append(words[0])
adv_imgs.append(words[1])
# labels.append(label)
self.clean_imgs = clean_imgs # 最主要就是要生成这个list， 然后DataLoader中给index，通过getitem读取图片数据
self.adv_imgs = adv_imgs
self.label = label
def __getitem__(self, index):
clean_address = self.clean_imgs[index]
adv_address = self.adv_imgs[index]
clean_img = TM.preprocess_image(cv2.imread(clean_address))
adv_img = TM.preprocess_image(cv2.imread(adv_address))
# print(clean_img.type)
clean_img = tf.to_tensor(clean_img)
adv_img = tf.to_tensor(adv_img)
return (adv_img - clean_img), self.label
def __len__(self):
return len(self.clean_imgs)
class Labeled_FaceDataset_incremental(Dataset):
def __init__(self, txt_path, label, known):
fh = open(txt_path, 'r')
clean_imgs = []
adv_imgs = []
# labels = []
for line in fh:
line = line.rstrip()
words = line.split()
clean_imgs.append(words[0])
adv_imgs.append(words[1])
# labels.append(label)
self.clean_imgs = clean_imgs # 最主要就是要生成这个list， 然后DataLoader中给index，通过getitem读取图片数据
self.adv_imgs = adv_imgs
self.label = label
self.known = known
def __getitem__(self, index):
clean_address = self.clean_imgs[index]
adv_address = self.adv_imgs[index]
clean_img = TM.preprocess_image(cv2.imread(clean_address))
adv_img = TM.preprocess_image(cv2.imread(adv_address))
# print(clean_img.type)
clean_img = tf.to_tensor(clean_img)
adv_img = tf.to_tensor(adv_img)
return (adv_img - clean_img), self.label, self.known
def __len__(self):
return len(self.clean_imgs)