-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathdata_utils.py
271 lines (215 loc) · 14.2 KB
/
data_utils.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
263
264
265
266
267
268
269
270
from datasets.glas import GLAS_Dataset
from datasets.kvasirseg import KVASIRSEG_Dataset
from datasets.polypgen import PolypGen_Dataset
from datasets.cityscapes import CITYSCAPES_Dataset
from datasets.pascal_voc import PASCAL_VOC_Dataset
from datasets.camvid import CAMVID_Dataset
from datasets.isic import Skin_Dataset
from datasets.refuge import Refuge_Dataset
from datasets.rite import RITE_Dataset
from datasets.endovis import Endovis_Dataset
from datasets.chestxdet import ChestXDet_Dataset
from torch.utils.data import DataLoader
import yaml
import torch
import matplotlib.pyplot as plt
from sklearn.decomposition import PCA
from sklearn.preprocessing import StandardScaler
import sys
# sys.path.append('../endovis17')
def get_data(data_config, center_num=1):
# print(data_config)
dataset_dict = {}
dataset_sizes = {}
dataloader_dict = {}
if data_config['name']=='GLAS':
dataset_dict['train'] = GLAS_Dataset(data_config, shuffle_list=True, is_train=True, apply_norm=data_config['use_norm'], no_text_mode=data_config['no_text_mode'])
dataset_dict['val'] = GLAS_Dataset(data_config, shuffle_list=False, apply_norm=data_config['use_norm'], is_train=False, no_text_mode=data_config['no_text_mode'])
dataset_dict['test'] = GLAS_Dataset(data_config, shuffle_list=False, is_train=False, is_test=True, apply_norm=data_config['use_norm'], no_text_mode=data_config['no_text_mode'])
dataset_sizes['train'] = len(dataset_dict['train'])
dataset_sizes['val'] = len(dataset_dict['val'])
dataset_sizes['test'] = len(dataset_dict['test'])
print(dataset_sizes)
# dataloader_dict['train'] = DataLoader(dataset_dict['train'], data_config['batch_size'], shuffle=True)
# dataloader_dict['val'] = DataLoader(dataset_dict['val'], data_config['batch_size'], shuffle=True)
elif data_config['name']=='KVASIRSEG':
dataset_dict['train'] = KVASIRSEG_Dataset(data_config, shuffle_list=True, is_train=True, apply_norm=data_config['use_norm'], no_text_mode=data_config['no_text_mode'])
dataset_dict['val'] = KVASIRSEG_Dataset(data_config, shuffle_list=False, apply_norm=data_config['use_norm'], is_train=False, no_text_mode=data_config['no_text_mode'])
dataset_dict['test'] = KVASIRSEG_Dataset(data_config, shuffle_list=False, is_train=False, is_test=True, apply_norm=data_config['use_norm'], no_text_mode=data_config['no_text_mode'])
dataset_sizes['train'] = len(dataset_dict['train'])
dataset_sizes['val'] = len(dataset_dict['val'])
dataset_sizes['test'] = len(dataset_dict['test'])
# dataloader_dict['train'] = DataLoader(dataset_dict['train'], data_config['batch_size'], shuffle=True)
# dataloader_dict['val'] = DataLoader(dataset_dict['val'], data_config['batch_size'], shuffle=True)
print(dataset_sizes)
elif data_config['name']=='POLYPGEN':
dataset_dict['train'] = PolypGen_Dataset(data_config, shuffle_list=True, is_train=True, apply_norm=data_config['use_norm'], center_num=center_num)
dataset_dict['val'] = PolypGen_Dataset(data_config, shuffle_list=False, apply_norm=data_config['use_norm'], is_train=False, center_num=center_num)
dataset_dict['test'] = PolypGen_Dataset(data_config, shuffle_list=False, is_train=False, is_test=True, apply_norm=data_config['use_norm'], center_num=center_num)
dataset_dict['name'] = str(center_num)
dataset_sizes['train'] = len(dataset_dict['train'])
dataset_sizes['val'] = len(dataset_dict['val'])
dataset_sizes['test'] = len(dataset_dict['test'])
# dataloader_dict['train'] = DataLoader(dataset_dict['train'], data_config['batch_size'], shuffle=True)
# dataloader_dict['val'] = DataLoader(dataset_dict['val'], data_config['batch_size'], shuffle=True)
# print(dataset_sizes)
elif data_config['name']=='SKIN':
dataset_dict['train'] = Skin_Dataset(data_config, shuffle_list=True, is_train=True, apply_norm=data_config['use_norm'], center_num=center_num)
dataset_dict['val'] = Skin_Dataset(data_config, shuffle_list=False, apply_norm=data_config['use_norm'], is_train=False, center_num=center_num)
dataset_dict['test'] = Skin_Dataset(data_config, shuffle_list=False, is_train=False, is_test=True, apply_norm=data_config['use_norm'], center_num = center_num)
dataset_dict['name'] = str(center_num)
dataset_sizes['train'] = len(dataset_dict['train'])
dataset_sizes['val'] = len(dataset_dict['val'])
dataset_sizes['test'] = len(dataset_dict['test'])
# dataloader_dict['train'] = DataLoader(dataset_dict['train'], data_config['batch_size'], shuffle=True)
# dataloader_dict['val'] = DataLoader(dataset_dict['val'], data_config['batch_size'], shuffle=True)
elif data_config['name']=='REFUGE':
dataset_dict['train'] = Refuge_Dataset(data_config, shuffle_list=True, is_train=True, apply_norm=data_config['use_norm'], no_text_mode=data_config['no_text_mode'])
dataset_dict['val'] = Refuge_Dataset(data_config, shuffle_list=False, apply_norm=data_config['use_norm'], is_train=False, no_text_mode=data_config['no_text_mode'])
dataset_dict['test'] = Refuge_Dataset(data_config, shuffle_list=False, is_train=False, is_test=True, apply_norm=data_config['use_norm'], no_text_mode=data_config['no_text_mode'])
dataset_sizes['train'] = len(dataset_dict['train'])
dataset_sizes['val'] = len(dataset_dict['val'])
dataset_sizes['test'] = len(dataset_dict['test'])
print(dataset_sizes)
# dataloader_dict['train'] = DataLoader(dataset_dict['train'], data_config['batch_size'], shuffle=True)
# dataloader_dict['val'] = DataLoader(dataset_dict['val'], data_config['batch_size'], shuffle=True)
elif data_config['name']=='RITE':
dataset_dict['train'] = RITE_Dataset(data_config, shuffle_list=True, is_train=True, apply_norm=data_config['use_norm'], no_text_mode=data_config['no_text_mode'])
dataset_dict['val'] = RITE_Dataset(data_config, shuffle_list=False, apply_norm=data_config['use_norm'], is_train=False, no_text_mode=data_config['no_text_mode'])
dataset_dict['test'] = RITE_Dataset(data_config, shuffle_list=False, is_train=False, is_test=True, apply_norm=data_config['use_norm'], no_text_mode=data_config['no_text_mode'])
dataset_sizes['train'] = len(dataset_dict['train'])
dataset_sizes['val'] = len(dataset_dict['val'])
dataset_sizes['test'] = len(dataset_dict['test'])
print(dataset_sizes)
# dataloader_dict['train'] = DataLoader(dataset_dict['train'], data_config['batch_size'], shuffle=True)
# dataloader_dict['val'] = DataLoader(dataset_dict['val'], data_config['batch_size'], shuffle=True)
elif data_config['name']=='ENDOVIS':
dataset_dict['train'] = Endovis_Dataset(data_config, shuffle_list=True, is_train=True, apply_norm=data_config['use_norm'], no_text_mode=data_config['no_text_mode'])
dataset_dict['val'] = Endovis_Dataset(data_config, shuffle_list=False, apply_norm=data_config['use_norm'], is_train=False, no_text_mode=data_config['no_text_mode'])
dataset_dict['test'] = Endovis_Dataset(data_config, shuffle_list=False, is_train=False, is_test=True, apply_norm=data_config['use_norm'], no_text_mode=data_config['no_text_mode'])
dataset_sizes['train'] = len(dataset_dict['train'])
dataset_sizes['val'] = len(dataset_dict['val'])
dataset_sizes['test'] = len(dataset_dict['test'])
print(dataset_sizes)
# dataloader_dict['train'] = DataLoader(dataset_dict['train'], data_config['batch_size'], shuffle=True)
# dataloader_dict['val'] = DataLoader(dataset_dict['val'], data_config['batch_size'], shuffle=True)
elif data_config['name']=='CHESTXDET':
dataset_dict['train'] = ChestXDet_Dataset(data_config, shuffle_list=True, is_train=True, apply_norm=data_config['use_norm'], no_text_mode=data_config['no_text_mode'])
dataset_dict['val'] = ChestXDet_Dataset(data_config, shuffle_list=False, apply_norm=data_config['use_norm'], is_train=False, no_text_mode=data_config['no_text_mode'])
dataset_dict['test'] = ChestXDet_Dataset(data_config, shuffle_list=False, is_train=False, is_test=True, apply_norm=data_config['use_norm'], no_text_mode=data_config['no_text_mode'])
dataset_sizes['train'] = len(dataset_dict['train'])
dataset_sizes['val'] = len(dataset_dict['val'])
dataset_sizes['test'] = len(dataset_dict['test'])
print(dataset_sizes)
elif data_config['name'] == 'CITYSCAPES':
dataset_dict['train'] = CITYSCAPES_Dataset(data_config, shuffle_list=True, is_train=True, apply_norm=data_config['use_norm'], center_num=center_num)
dataset_dict['val'] = CITYSCAPES_Dataset(data_config, shuffle_list=False, apply_norm=data_config['use_norm'], is_train=False, center_num=center_num)
dataset_dict['test'] = CITYSCAPES_Dataset(data_config, shuffle_list=False, is_train=False, is_test=True, apply_norm=data_config['use_norm'], center_num=center_num)
dataset_dict['pure_test'] = CITYSCAPES_Dataset(data_config, shuffle_list=False, is_train=False, is_test=True, apply_norm=data_config['use_norm'], center_num=center_num, pure_test=True)
dataset_dict['name'] = str(center_num)
dataset_sizes['train'] = len(dataset_dict['train'])
dataset_sizes['val'] = len(dataset_dict['val'])
dataset_sizes['test'] = len(dataset_dict['test'])
dataset_sizes['pure_test'] = len(dataset_dict['pure_test'])
elif data_config['name'] == 'CAMVID':
dataset_dict['train'] = CAMVID_Dataset(data_config, shuffle_list=True, is_train=True, apply_norm=data_config['use_norm'], center_num=center_num)
dataset_dict['val'] = CAMVID_Dataset(data_config, shuffle_list=False, apply_norm=data_config['use_norm'], is_train=False, center_num=center_num)
dataset_dict['test'] = CAMVID_Dataset(data_config, shuffle_list=False, is_train=False, is_test=True, apply_norm=data_config['use_norm'], center_num=center_num)
dataset_dict['name'] = str(center_num)
dataset_sizes['train'] = len(dataset_dict['train'])
dataset_sizes['val'] = len(dataset_dict['val'])
dataset_sizes['test'] = len(dataset_dict['test'])
elif data_config['name'] == 'PASCAL_VOC':
dataset_dict['train'] = PASCAL_VOC_Dataset(data_config, shuffle_list=True, is_train=True, apply_norm=data_config['use_norm'], center_num=center_num)
dataset_dict['val'] = PASCAL_VOC_Dataset(data_config, shuffle_list=False, apply_norm=data_config['use_norm'], is_train=False, center_num=center_num)
dataset_dict['test'] = PASCAL_VOC_Dataset(data_config, shuffle_list=False, is_train=False, is_test=True, apply_norm=data_config['use_norm'], center_num=center_num)
dataset_dict['pure_test'] = PASCAL_VOC_Dataset(data_config, shuffle_list=False, is_train=False, is_test=True, apply_norm=data_config['use_norm'], center_num=center_num, pure_test=True)
dataset_dict['name'] = str(center_num)
dataset_sizes['train'] = len(dataset_dict['train'])
dataset_sizes['val'] = len(dataset_dict['val'])
dataset_sizes['test'] = len(dataset_dict['test'])
dataset_sizes['pure_test'] = len(dataset_dict['pure_test'])
return dataset_dict
def visualize_PCA(dataset_dict_super, dataset_dicts, phase='train'):
images_centers = []
c = ['red', 'green', 'blue', 'black', 'yellow', 'cyan', 'orange', 'brown', 'pink', 'gray']
data_super = dataset_dict_super[phase]
images = []
count = 0
for img,_,_,_ in data_super:
count+=1
# if count>5:
# continue
images.append(img)
l = len(images)
images = torch.stack(images,dim=0).permute(0,2,3,1).cpu().numpy().reshape((l,-1))
# Scale data before applying PCA
scaling=StandardScaler()
# Use fit and transform method
scaling.fit(images)
Scaled_data=scaling.transform(images)
# Set the n_components=3
principal=PCA(n_components=2)
principal.fit(Scaled_data)
for i in range(len(dataset_dicts)):
data = dataset_dicts[i][phase]
images = []
count = 0
for img,_,_,_ in data:
count+=1
# if count>5:
# continue
images.append(img)
l = len(images)
images = torch.stack(images,dim=0).permute(0,2,3,1).cpu().numpy().reshape((l,-1))
x=principal.transform(images)
images_centers.append(x)
#visualization
plt.scatter(x=x[:,0], y=x[:,1],c=c[i])
plt.legend(['C'+str(i+1) for i in range(len(dataset_dicts))])
plt.savefig("./tmp2.png")
if __name__ == '__main__':
with open(sys.argv[1], 'r') as f:
config = yaml.load(f, Loader=yaml.FullLoader)
#1
# dataset_dict = get_data(config, center_num=1)
# print(dataset_dict['train'][0][0].shape)
# print(dataset_dict['train'][0][1].shape)
# print(dataset_dict['train'][52][1].any())
# plt.imshow(dataset_dict['train'][0][0].permute(1,2,0))
# plt.show()
# plt.imshow(dataset_dict['train'][0][1], cmap='gray')
# plt.show()
#2
super_dataset_dict = get_data(config, center_num='super')
dataset_dicts = [get_data(config, center_num=i) for i in range(1,19)]
# dataset_dicts = [get_data(config, center_num=i) for i in [1, 2, 3]]
print("Training stats: ")
for i in range(len(dataset_dicts)):
l = len(dataset_dicts[i]['train'])
print(f'data center {i+1} has {l} training points')
print("Val stats: ")
for i in range(len(dataset_dicts)):
l = len(dataset_dicts[i]['val'])
print(f'data center {i+1} has {l} training points')
print("Testing stats: ")
for i in range(len(dataset_dicts)):
l = len(dataset_dicts[i]['test'])
print(f'data center {i+1} has {l} training points')
# #3
# visualize_PCA(super_dataset_dict, dataset_dicts)
#4
#visualize images
# config['use_norm'] = False
# dd = get_data(config, center_num='super')
# print(len(dd['train']))
# print(len(dd['val']))
# ridx = 10
# torch.save(dd['val'][ridx][0],'tmp3.pt')
# plt.imshow(dd['val'][ridx][0].permute(1,2,0))
# plt.savefig('tmp3.png')
# label = dd['val'][ridx][1]
# print(label.shape)
# print(label.max())
# plt.imshow(label.permute(1,2,0)[:,:,16], cmap='gray')
# plt.savefig('tmp4.png')