-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathsvhn_loader.py
123 lines (105 loc) · 5.12 KB
/
svhn_loader.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
import torch.utils.data as data
from PIL import Image
import os
import os.path
import numpy as np
class SVHN(data.Dataset):
url = ""
filename = ""
file_md5 = ""
split_list = {
'train': ["http://ufldl.stanford.edu/housenumbers/train_32x32.mat",
"train_32x32.mat", "e26dedcc434d2e4c54c9b2d4a06d8373"],
'test': ["http://ufldl.stanford.edu/housenumbers/test_32x32.mat",
"test_32x32.mat", "eb5a983be6a315427106f1b164d9cef3"],
'extra': ["http://ufldl.stanford.edu/housenumbers/extra_32x32.mat",
"extra_32x32.mat", "a93ce644f1a588dc4d68dda5feec44a7"],
'train_and_extra': [
["http://ufldl.stanford.edu/housenumbers/train_32x32.mat",
"train_32x32.mat", "e26dedcc434d2e4c54c9b2d4a06d8373"],
["http://ufldl.stanford.edu/housenumbers/extra_32x32.mat",
"extra_32x32.mat", "a93ce644f1a588dc4d68dda5feec44a7"]]}
def __init__(self, root, split='train',
transform=None, target_transform=None, download=False):
self.root = root
self.transform = transform
self.target_transform = target_transform
self.split = split # training set or test set or extra set
if self.split not in self.split_list:
raise ValueError('Wrong split entered! Please use split="train" '
'or split="extra" or split="test" '
'or split="train_and_extra" ')
if self.split == "train_and_extra":
self.url = self.split_list[split][0][0]
self.filename = self.split_list[split][0][1]
self.file_md5 = self.split_list[split][0][2]
else:
self.url = self.split_list[split][0]
self.filename = self.split_list[split][1]
self.file_md5 = self.split_list[split][2]
# import here rather than at top of file because this is
# an optional dependency for torchvision
import scipy.io as sio
# reading(loading) mat file as array
loaded_mat = sio.loadmat(os.path.join(root, self.filename))
if self.split == "test":
self.data = loaded_mat['X']
self.targets = loaded_mat['y']
# Note label 10 == 0 so modulo operator required
self.targets = (self.targets % 10).squeeze() # convert to zero-based indexing
self.data = np.transpose(self.data, (3, 2, 0, 1))
else:
self.data = loaded_mat['X']
self.targets = loaded_mat['y']
if self.split == "train_and_extra":
extra_filename = self.split_list[split][1][1]
loaded_mat = sio.loadmat(os.path.join(root, extra_filename))
self.data = np.concatenate([self.data,
loaded_mat['X']], axis=3)
self.targets = np.vstack((self.targets,
loaded_mat['y']))
# Note label 10 == 0 so modulo operator required
self.targets = (self.targets % 10).squeeze() # convert to zero-based indexing
self.data = np.transpose(self.data, (3, 2, 0, 1))
def __getitem__(self, index):
if self.split == "test":
img, target = self.data[index], self.targets[index]
else:
img, target = self.data[index], self.targets[index]
# doing this so that it is consistent with all other datasets
# to return a PIL Image
img = Image.fromarray(np.transpose(img, (1, 2, 0)))
if self.transform is not None:
img = self.transform(img)
if self.target_transform is not None:
target = self.target_transform(target)
return img, target
def __len__(self):
if self.split == "test":
return len(self.data)
else:
return len(self.data)
def _check_integrity(self):
root = self.root
if self.split == "train_and_extra":
md5 = self.split_list[self.split][0][2]
fpath = os.path.join(root, self.filename)
train_integrity = check_integrity(fpath, md5)
extra_filename = self.split_list[self.split][1][1]
md5 = self.split_list[self.split][1][2]
fpath = os.path.join(root, extra_filename)
return check_integrity(fpath, md5) and train_integrity
else:
md5 = self.split_list[self.split][2]
fpath = os.path.join(root, self.filename)
return check_integrity(fpath, md5)
def download(self):
if self.split == "train_and_extra":
md5 = self.split_list[self.split][0][2]
download_url(self.url, self.root, self.filename, md5)
extra_filename = self.split_list[self.split][1][1]
md5 = self.split_list[self.split][1][2]
download_url(self.url, self.root, extra_filename, md5)
else:
md5 = self.split_list[self.split][2]
download_url(self.url, self.root, self.filename, md5)