-
Notifications
You must be signed in to change notification settings - Fork 30
/
imagenet.py
166 lines (139 loc) · 6.41 KB
/
imagenet.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
"""ReBias
Copyright (c) 2020-present NAVER Corp.
MIT license
9-Class ImageNet wrapper. Many codes are borrowed from the official torchvision dataset.
https://github.com/pytorch/vision/blob/master/torchvision/datasets/imagenet.py
The following nine classes are selected to build the subset:
dog, cat, frog, turtle, bird, monkey, fish, crab, insect
"""
import os
from PIL import Image
from torchvision import transforms
import torch
import torch.utils.data
IMG_EXTENSIONS = [
'.jpg', '.JPG', '.jpeg', '.JPEG',
'.png', '.PNG', '.ppm', '.PPM', '.bmp', '.BMP',
]
CLASS_TO_INDEX = {'n01641577': 2, 'n01644373': 2, 'n01644900': 2, 'n01664065': 3, 'n01665541': 3,
'n01667114': 3, 'n01667778': 3, 'n01669191': 3, 'n01819313': 4, 'n01820546': 4,
'n01833805': 4, 'n01843383': 4, 'n01847000': 4, 'n01978287': 7, 'n01978455': 7,
'n01980166': 7, 'n01981276': 7, 'n02085620': 0, 'n02099601': 0, 'n02106550': 0,
'n02106662': 0, 'n02110958': 0, 'n02123045': 1, 'n02123159': 1, 'n02123394': 1,
'n02123597': 1, 'n02124075': 1, 'n02174001': 8, 'n02177972': 8, 'n02190166': 8,
'n02206856': 8, 'n02219486': 8, 'n02486410': 5, 'n02487347': 5, 'n02488291': 5,
'n02488702': 5, 'n02492035': 5, 'n02607072': 6, 'n02640242': 6, 'n02641379': 6,
'n02643566': 6, 'n02655020': 6}
def is_image_file(filename):
return any(filename.endswith(extension) for extension in IMG_EXTENSIONS)
def make_dataset(dir, class_to_idx, data='ImageNet'):
# dog, cat, frog, turtle, bird, monkey, fish, crab, insect
RESTRICTED_RANGES = [(151, 254), (281, 285), (30, 32), (33, 37), (89, 97),
(372, 378), (393, 397), (118, 121), (306, 310)]
range_sets = [set(range(s, e + 1)) for s, e in RESTRICTED_RANGES]
class_to_idx_ = {}
if data == 'ImageNet-A':
for class_name, idx in class_to_idx.items():
try:
class_to_idx_[class_name] = CLASS_TO_INDEX[class_name]
except Exception:
pass
elif data == 'ImageNet-C':
# TODO
pass
else: # ImageNet
for class_name, idx in class_to_idx.items():
for new_idx, range_set in enumerate(range_sets):
if idx in range_set:
if new_idx == 0: # classes that overlap with ImageNet-A
if idx in [151, 207, 234, 235, 254]:
class_to_idx_[class_name] = new_idx
elif new_idx == 4:
if idx in [89, 90, 94, 96, 97]:
class_to_idx_[class_name] = new_idx
elif new_idx == 5:
if idx in [372, 373, 374, 375, 378]:
class_to_idx_[class_name] = new_idx
else:
class_to_idx_[class_name] = new_idx
images = []
dir = os.path.expanduser(dir)
a = sorted(class_to_idx_.keys())
for target in a:
d = os.path.join(dir, target)
if not os.path.isdir(d):
continue
for root, _, fnames in sorted(os.walk(d)):
for fname in fnames:
if is_image_file(fname):
path = os.path.join(root, fname)
item = (path, class_to_idx_[target])
images.append(item)
return images, class_to_idx_
def find_classes(dir):
classes = [d for d in os.listdir(dir) if os.path.isdir(os.path.join(dir, d))]
classes.sort()
class_to_idx = {classes[i]: i for i in range(len(classes))}
return classes, class_to_idx
def pil_loader(path):
with open(path, 'rb') as f:
with Image.open(f) as img:
return img.convert('RGB')
class ImageFolder(torch.utils.data.Dataset):
def __init__(self, root, transform=None, target_transform=None, loader=pil_loader,
train=True, val_data='ImageNet'):
classes, class_to_idx = find_classes(root)
imgs, class_to_idx_ = make_dataset(root, class_to_idx, val_data)
if len(imgs) == 0:
raise (RuntimeError("Found 0 images in subfolders of: " + root + "\n"
"Supported image extensions are: " + ",".join(
IMG_EXTENSIONS)))
self.root = root
self.dataset = imgs
self.classes = classes
self.class_to_idx = class_to_idx_
self.transform = transform
self.target_transform = target_transform
self.loader = loader
self.train = train
self.val_data = val_data
self.clusters = []
for i in range(3):
self.clusters.append(torch.load('clusters/cluster_label_{}.pth'.format(i+1)))
def __getitem__(self, index):
path, target = self.dataset[index]
img = self.loader(path)
if self.transform is not None:
img = self.transform(img)
if self.target_transform is not None:
target = self.target_transform(target)
if not self.train and self.val_data == 'ImageNet':
bias_target = [self.clusters[0][index],
self.clusters[1][index],
self.clusters[2][index]]
return img, target, bias_target
else:
return img, target, target
def __len__(self):
return len(self.dataset)
def get_imagenet_dataloader(root, batch_size, train=True, num_workers=8,
load_size=256, image_size=224, val_data='ImageNet'):
if train:
transform = transforms.Compose([
transforms.RandomResizedCrop(image_size),
transforms.RandomHorizontalFlip(),
transforms.ToTensor(),
transforms.Normalize(mean=(0.485, 0.456, 0.406), std=(0.229, 0.224, 0.225))])
else:
transform = transforms.Compose([
transforms.Resize(load_size),
transforms.CenterCrop(image_size),
transforms.ToTensor(),
transforms.Normalize(mean=(0.485, 0.456, 0.406), std=(0.229, 0.224, 0.225))])
dataset = ImageFolder(root, transform=transform, train=train, val_data=val_data)
dataloader = torch.utils.data.DataLoader(dataset=dataset,
batch_size=batch_size,
shuffle=train,
num_workers=num_workers,
pin_memory=True)
return dataloader