-
Notifications
You must be signed in to change notification settings - Fork 6
/
Copy pathbanknoteDataLoader.py
216 lines (181 loc) · 11.6 KB
/
banknoteDataLoader.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
import cv2
import numpy as np
import torch.utils.data
from tqdm import tqdm
import math
import psutil
import gc
import torch
from torch.autograd import Variable
import torchvision.transforms as transforms
import torchvision.models as models
from models.customResnet50 import CustomResNet50
import pdb
from option import Options
from dataset.banknote_pytorch import FullBanknotePairs, FullBanknote, FullBanknoteOneShot, FullBanknoteTriplets
#torch.utils.data.dataloader.default_collate = (lambda default_collate = torch.utils.data.dataloader.default_collate: \
# lambda batch: batch if all(map(torch.is_tensor, batch)) \
# and any([tensor.size() != batch[0].size() for tensor in batch]) else default_collate(batch))()
class banknoteDataLoader():
def __init__(self,type=FullBanknotePairs, opt=Options().parse(), fcn = None, train_mean=None, train_std=None):
self.type = type
self.opt = opt
self.fcn = fcn
if train_mean is None and train_std is None:
self.train_mean = None
self.train_std = None
else:
self.train_mean = train_mean
self.train_std = train_std
def getlstTransforms(self, train = 'train'):
lst_transforms = []
if not(self.opt.imageSize is None):
lst_transforms.append(transforms.Resize((self.opt.imageSize,self.opt.imageSize)))
if train == 'train':
lst_transforms.append(transforms.ColorJitter(brightness=0.4,contrast=0.4,saturation=0.4))
lst_transforms.append(transforms.RandomAffine(degrees=(0,10), translate=(0.1, 0.1), scale=(0.8, 1.2)))
if self.opt.imageSize is None:
lst_transforms.append(transforms.RandomCrop(size=224))
if not(train == 'train'):
#lst_transforms.append(transforms.CenterCrop(size=224)) # fixed zone
lst_transforms.append(transforms.RandomCrop(size=224)) # random zone
lst_transforms.append(transforms.ToTensor())
if not(self.train_mean is None) and not(self.train_std is None):
lst_transforms.append(transforms.Normalize(torch.from_numpy(self.train_mean),torch.from_numpy(self.train_std)))
if self.opt.fcn_applyOnDataLoader:
lst_transforms.append(transforms.Lambda(lambda x: x.unsqueeze(0)))
if self.opt.cuda:
lst_transforms.append(transforms.Lambda(lambda x: Variable(x.cuda(), requires_grad=True)))
else:
lst_transforms.append(transforms.Lambda(lambda x: Variable(x, requires_grad=True)))
lst_transforms.append(transforms.Lambda(lambda x: self.fcn(x)))
return lst_transforms
def get_mean_std(self):
lst_transforms = []
if not(self.opt.imageSize is None):
lst_transforms.append(transforms.Resize((self.opt.imageSize,self.opt.imageSize)))
lst_transforms.append(transforms.ColorJitter(brightness=0.4,contrast=0.4,saturation=0.4))
lst_transforms.append(transforms.RandomAffine(degrees=(0,10), translate=(0.1, 0.1), scale=(0.8, 1.2)))
if self.opt.imageSize is None:
lst_transforms.append(transforms.RandomCrop(size=224))
lst_transforms.append(transforms.ToTensor())
train_transform = transforms.Compose(lst_transforms)
kwargs = {'num_workers': self.opt.nthread, 'pin_memory': True} if self.opt.cuda else {}
if self.type == FullBanknotePairs or self.type == FullBanknote or self.type == FullBanknoteTriplets:
train_loader_mean_std = torch.utils.data.DataLoader(
FullBanknote(setType=self.opt.setType, root=self.opt.dataroot, train='train', size = self.opt.imageSize,
mode = 'generator_processor', path_tmp_data = self.opt.path_tmp_data,
transform=train_transform, target_transform=None),
#batch_size=self.opt.batchSize, shuffle=True, collate_fn = torch.utils.data.dataloader.default_collate, **kwargs)
batch_size=self.opt.batchSize, shuffle=True, **kwargs)
print('Calculate mean and std for training set....')
pbar = tqdm(enumerate(train_loader_mean_std))
tmp = []
for batch_idx, (data, labels) in pbar:
tmp.append(data.data.cpu().numpy())
pbar.set_description(
'[{}/{} ({:.0f}%)]\t'.format(
batch_idx * len(data), len(train_loader_mean_std.dataset),
100. * batch_idx / len(train_loader_mean_std)))
# Memory problems if we acumulate the images.
free_mem = psutil.virtual_memory().available / (1024.0 ** 3)
# print('Free mem: %f' % free_mem)
if int(math.floor(free_mem)) == 0:
break
tmp = np.vstack(tmp)
train_mean = tmp.mean(axis=0)
train_std = tmp.std(axis=0)
# Free memory
tmp = []
data = []
labels = []
train_loader_mean_std = None
gc.collect()
return train_mean, train_std
def get(self, dataPartition = ['train','val','test'] ,rnd_seed = 42):
#if self.train_mean is None and self.train_std is None and not(self.opt.imageSize is None):
if self.train_mean is None and self.train_std is None:
train_mean, train_std = self.get_mean_std()
self.train_mean = train_mean
self.train_std = train_std
#kwargs = {'num_workers': self.opt.nthread, 'pin_memory': True} if self.opt.cuda else {}
if self.opt.cuda:
kwargs = {'num_workers': self.opt.nthread, 'pin_memory': True}
else:
kwargs = {'num_workers': self.opt.nthread}
#kwargs = {}
train_transform = transforms.Compose(self.getlstTransforms(train = 'train'))
if len(dataPartition) > 0 and not(dataPartition[0] is None):
if self.type == FullBanknote:
datasetParams = self.type(setType=self.opt.setType, root=self.opt.dataroot, train=dataPartition[0],
size = self.opt.imageSize,
mode = self.opt.mode, path_tmp_data = self.opt.path_tmp_data,
transform=train_transform, target_transform=None)
elif self.type == FullBanknotePairs or self.type == FullBanknoteTriplets:
datasetParams = self.type(setType=self.opt.setType, root=self.opt.dataroot, train=dataPartition[0],
size = self.opt.imageSize, numTrials=self.opt.batchSize,
mode = self.opt.mode, path_tmp_data = self.opt.path_tmp_data,
transform=train_transform, target_transform=None)
elif self.type == FullBanknoteOneShot:
datasetParams = self.type(setType=self.opt.setType, root=self.opt.dataroot, train=dataPartition[0],
size = self.opt.imageSize,
transform=train_transform, target_transform=None,
mode = self.opt.mode, path_tmp_data = self.opt.path_tmp_data,
sameClass = self.opt.datasetBanknoteOneShotSameClass,
n_way = self.opt.one_shot_n_way, n_shot = self.opt.one_shot_n_shot,
numTrials=self.opt.batchSize)
train_loader = torch.utils.data.DataLoader(
datasetParams,
batch_size=self.opt.batchSize, shuffle=True, **kwargs)
else:
train_loader = None
eval_test_transform = transforms.Compose(self.getlstTransforms(train = 'val_test'))
if len(dataPartition) > 0 and not(dataPartition[1] is None):
if self.type == FullBanknote:
datasetParams = self.type(setType=self.opt.setType, root=self.opt.dataroot, train=dataPartition[1],
size = self.opt.imageSize,
mode = self.opt.mode, path_tmp_data = self.opt.path_tmp_data,
transform=eval_test_transform, target_transform=None)
elif self.type == FullBanknotePairs or self.type == FullBanknoteTriplets:
datasetParams = self.type(setType=self.opt.setType, root=self.opt.dataroot, train=dataPartition[1],
size = self.opt.imageSize, numTrials=self.opt.batchSize,
mode = self.opt.mode, path_tmp_data = self.opt.path_tmp_data,
transform=eval_test_transform, target_transform=None)
elif self.type == FullBanknoteOneShot:
datasetParams = self.type(setType=self.opt.setType, root=self.opt.dataroot, train=dataPartition[1],
size = self.opt.imageSize,
transform=train_transform, target_transform=None,
mode = self.opt.mode, path_tmp_data = self.opt.path_tmp_data,
sameClass = self.opt.datasetBanknoteOneShotSameClass,
n_way = self.opt.one_shot_n_way, n_shot = self.opt.one_shot_n_shot,
numTrials=self.opt.batchSize)
val_loader = torch.utils.data.DataLoader(
datasetParams,
batch_size=self.opt.batchSize, shuffle=True, **kwargs)
else:
val_loader = None
if len(dataPartition) > 0 and not(dataPartition[2] is None):
if self.type == FullBanknote:
datasetParams = self.type(setType=self.opt.setType, root=self.opt.dataroot, train=dataPartition[2],
size = self.opt.imageSize,
mode = self.opt.mode, path_tmp_data = self.opt.path_tmp_data,
transform=eval_test_transform, target_transform=None)
elif self.type == FullBanknotePairs or self.type == FullBanknoteTriplets:
datasetParams = self.type(setType=self.opt.setType, root=self.opt.dataroot, train=dataPartition[2],
size = self.opt.imageSize, numTrials=self.opt.batchSize,
mode = self.opt.mode, path_tmp_data = self.opt.path_tmp_data,
transform=eval_test_transform, target_transform=None)
elif self.type == FullBanknoteOneShot:
datasetParams = self.type(setType=self.opt.setType, root=self.opt.dataroot, train=dataPartition[2],
size = self.opt.imageSize,
transform=train_transform, target_transform=None,
mode = self.opt.mode, path_tmp_data = self.opt.path_tmp_data,
sameClass = self.opt.datasetBanknoteOneShotSameClass,
n_way = self.opt.one_shot_n_way, n_shot = self.opt.one_shot_n_shot,
numTrials=self.opt.batchSize)
test_loader = torch.utils.data.DataLoader(
datasetParams,
batch_size=self.opt.batchSize, shuffle=True, **kwargs)
else:
test_loader = None
return train_loader, val_loader, test_loader