-
Notifications
You must be signed in to change notification settings - Fork 0
/
dataset.py
executable file
·215 lines (168 loc) · 8.59 KB
/
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
import csv
import os
import pickle
import torch
from PIL import Image
from torch.utils.data import Dataset
from torchvision import transforms
from tqdm import tqdm
DEFAULT_TRANSFORM = transforms.Compose([
transforms.Resize(size=256, interpolation=transforms.InterpolationMode.BICUBIC),
transforms.CenterCrop((224, 224)),
transforms.RandomHorizontalFlip(p=0.5),
transforms.ToTensor(),
transforms.Normalize(mean=[0.48145466, 0.4578275, 0.40821073],
std=[0.26862954, 0.26130258, 0.27577711]),
])
# PIAA Dataset
class FlickrAESPIAADataset(Dataset):
def __init__(self, root_dir, label_file, image_size, transform=None,
split='train', worker='', train_list=None, ):
self.root_dir = root_dir
# This might throw an error in a ddp setting. Simply run the script again.
if not os.path.isfile(os.path.join(root_dir, f"{label_file.split('.')[0]}_metadata.pkl")):
with open(os.path.join(root_dir, label_file), 'r') as f:
reader = csv.reader(f)
lines = [i for i in reader]
self.data_list = [{'worker':line[0], 'MOS':float(line[2]), 'image_name':line[1]} for line in lines[1:]]
with open(os.path.join(root_dir, f"{label_file.split('.')[0]}_metadata.pkl"), 'wb') as f:
pickle.dump(self.data_list, f)
else:
with open(os.path.join(root_dir, f"{label_file.split('.')[0]}_metadata.pkl"), 'rb') as f:
self.data_list = pickle.load(f)
train_list = [i for i in self.data_list if i['image_name'] in train_list and i['worker'] == worker]
if split == 'train':
self.data_list = train_list
elif split == 'test':
self.data_list = [i for i in self.data_list if i not in train_list and i['worker'] == worker]
print(f'Flickr AES dataset for worker: {worker}')
print(f'Flickr AES dataset size: {len(self.data_list)}')
self.transform = transform if transform is not None else DEFAULT_TRANSFORM
def __len__(self):
return len(self.data_list)
def __getitem__(self, idx):
data = {'image': None, 'MOS': None, 'image_name': '', 'worker': ''}
meta = self.data_list[idx]
image = Image.open(os.path.join(self.root_dir, '40K', meta['image_name']))
image = image.convert('RGB')
image = self.transform(image)
data['image'] = image
data['image_name'] = meta['image_name']
data['MOS'] = float(meta['MOS'])
data['worker'] = meta['worker']
return data
class REALCURDataset(Dataset):
def __init__(self, root_dir, label_file, image_size, transform=None,
split='train', worker='', train_list=None, ):
self.root_dir = root_dir
# This might throw an error in a ddp setting. Simply run the script again.
if not os.path.isfile(os.path.join(root_dir, f"metadata.pkl")):
with open(os.path.join(root_dir, label_file), 'r') as f:
reader = csv.reader(f)
lines = [i for i in reader]
self.data_list = [{'worker':line[0], 'MOS':float(line[2]), 'image_name':line[1]} for line in lines[1:]]
with open(os.path.join(root_dir, "metadata.pkl"), 'wb') as f:
pickle.dump(self.data_list, f)
else:
with open(os.path.join(root_dir, "metadata.pkl"), 'rb') as f:
self.data_list = pickle.load(f)
self.transform = transform if transform is not None else DEFAULT_TRANSFORM
self.worker = worker
train_list = [i for i in self.data_list if i['image_name'] in train_list and i['worker'] == worker]
if split == 'train':
self.data_list = train_list
elif split == 'test':
self.data_list = [i for i in self.data_list if i not in train_list and i['worker'] == worker]
print(f'REAL-CUR dataset for worker: {worker}')
print(f'REAL-CUR dataset size: {len(self.data_list)}')
def __len__(self):
return len(self.data_list)
def __getitem__(self, idx):
data = {'image': None, 'MOS': None, 'image_name': '', 'worker': ''}
meta = self.data_list[idx]
# data is from the Flickr AES dataset
image = Image.open(os.path.join(self.root_dir, 'images', self.worker, str(int(meta['MOS'])), meta['image_name']))
image = image.convert('RGB')
image = self.transform(image)
data['image'] = image
data['image_name'] = meta['image_name']
data['MOS'] = float(meta['MOS'])
data['worker'] = meta['worker']
return data
class PARAPIAADataset(Dataset):
def __init__(self, root_dir, label_file, image_size, transform=None,
split='train', worker='', train_list=None, ):
self.root_dir = root_dir
# This might throw an error in a ddp setting. Simply run the script again.
if not os.path.isfile(os.path.join(root_dir, f"metadata.pkl")):
with open(os.path.join(root_dir, label_file), 'r') as f:
reader = csv.reader(f)
lines = [i for i in reader]
self.data_list = [{'worker':line[0], 'MOS':2 * float(line[2]),
'image_name':line[1], 'session_id': line[3]} for line in lines[1:]]
with open(os.path.join(root_dir, "metadata.pkl"), 'wb') as f:
pickle.dump(self.data_list, f)
else:
with open(os.path.join(root_dir, "metadata.pkl"), 'rb') as f:
self.data_list = pickle.load(f)
self.transform = transform if transform is not None else DEFAULT_TRANSFORM
self.worker = worker
train_list = [i for i in self.data_list if i['image_name'] in train_list and i['worker'] == worker]
if split == 'train':
self.data_list = train_list
elif split == 'test':
self.data_list = [i for i in self.data_list if i not in train_list and i['worker'] == worker]
print(f'PARA PIAA dataset for worker: {worker}')
print(f'PARA PIAA dataset size: {len(self.data_list)}')
def __len__(self):
return len(self.data_list)
def __getitem__(self, idx):
data = {'image': None, 'MOS': None, 'image_name': '', 'worker': ''}
meta = self.data_list[idx]
# data is from the Flickr AES dataset
image = Image.open(os.path.join(self.root_dir, 'imgs_40', meta['session_id'], meta['image_name']))
image = image.convert('RGB')
image = self.transform(image)
data['image'] = image
data['image_name'] = meta['image_name']
data['MOS'] = float(meta['MOS'])
data['worker'] = meta['worker']
return data
class AADBDataset(Dataset):
def __init__(self, root_dir, label_file, image_size, transform=None,
split='train', worker='', train_list=None, ):
self.root_dir = root_dir
# This might throw an error in a ddp setting. Simply run the script again.
if not os.path.isfile(os.path.join(root_dir, f"metadata.pkl")):
with open(os.path.join(root_dir, label_file), 'r') as f:
reader = csv.reader(f)
lines = [i for i in reader]
self.data_list = [{'worker':line[0], 'MOS':float(line[2]), 'image_name':line[1]} for line in lines[1:]]
with open(os.path.join(root_dir, "metadata.pkl"), 'wb') as f:
pickle.dump(self.data_list, f)
else:
with open(os.path.join(root_dir, "metadata.pkl"), 'rb') as f:
self.data_list = pickle.load(f)
self.transform = transform if transform is not None else DEFAULT_TRANSFORM
self.worker = worker
train_list = [i for i in self.data_list if i['image_name'] in train_list and i['worker'] == worker]
if split == 'train':
self.data_list = train_list
elif split == 'test':
self.data_list = [i for i in self.data_list if i not in train_list and i['worker'] == worker]
print(f'AADB dataset for worker: {worker}')
print(f'AADB dataset size: {len(self.data_list)}')
def __len__(self):
return len(self.data_list)
def __getitem__(self, idx):
data = {'image': None, 'MOS': None, 'image_name': '', 'worker': ''}
meta = self.data_list[idx]
# data is from the Flickr AES dataset
image = Image.open(os.path.join(self.root_dir, 'datasetImages_originalSize', meta['image_name']))
image = image.convert('RGB')
image = self.transform(image)
data['image'] = image
data['image_name'] = meta['image_name']
data['MOS'] = float(meta['MOS'])
data['worker'] = meta['worker']
return data