-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathutils.py
155 lines (130 loc) · 5.84 KB
/
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
import cv2, os
from tqdm import tqdm
import itertools, imageio, torch, random
import matplotlib.pyplot as plt
import numpy as np
import torch.nn as nn
from torchvision import datasets
from scipy.misc import imresize
from torch.autograd import Variable
def retrieve_permutations(grid_size, classes):
all_perm = np.load('./permutations/permutations_hamming_max_%d_%d.npy' % (grid_size**2, classes))
# from range [1,9] to [0,8]
if all_perm.min() == 1:
all_perm = all_perm - 1
return all_perm
def read_flows(grid_size, classes):
all_perm = np.load('./permutations/flows_32_max_%d_%d.npy' % (grid_size**2, classes))
return all_perm
def data_load(path, subfolder, transform, batch_size, shuffle=False, drop_last=True):
dset = datasets.ImageFolder(path, transform)
ind = dset.class_to_idx[subfolder]
n = 0
for i in range(dset.__len__()):
if ind != dset.imgs[n][1]:
del dset.imgs[n]
n -= 1
n += 1
return torch.utils.data.DataLoader(dset, batch_size=batch_size, shuffle=shuffle, drop_last=drop_last)
def print_network(net):
num_params = 0
for param in net.parameters():
num_params += param.numel()
print(net)
print('Total number of parameters: %d' % num_params)
def initialize_weights(net):
for m in net.modules():
if isinstance(m, nn.Conv2d):
m.weight.data.normal_(0, 0.02)
m.bias.data.zero_()
elif isinstance(m, nn.ConvTranspose2d):
m.weight.data.normal_(0, 0.02)
m.bias.data.zero_()
elif isinstance(m, nn.Linear):
m.weight.data.normal_(0, 0.02)
m.bias.data.zero_()
elif isinstance(m, nn.BatchNorm2d):
m.weight.data.fill_(1)
m.bias.data.zero_()
def random_input4_combine(root, save):
file_list = os.listdir(root)
if not os.path.isdir(save):
os.makedirs(save)
n = 1
for f in tqdm(file_list):
order_original = [[0, 0], [0, 1], [1, 0], [1, 1]]
rgb_img = cv2.imread(os.path.join(root, f))
print('name:%s' % f)
rgb_img = cv2.resize(rgb_img, (256, 256))
random.shuffle(order_original)
# order_random = random.shuffle(order_original)
random_rgb_img = np.zeros([256,256,3], np.uint8)
random_rgb_img[0:128, 0:128] = rgb_img[order_original[0][0]*128:(order_original[0][0]+1)*128, order_original[0][1]*128:(order_original[0][1]+1)*128]
random_rgb_img[128:256, 0:128] = rgb_img[order_original[1][0]*128:(order_original[1][0]+1)*128, order_original[1][1]*128:(order_original[1][1]+1)*128]
random_rgb_img[0:128, 128:256] = rgb_img[order_original[2][0] * 128:(order_original[2][0]+1)*128, order_original[2][1] * 128:(order_original[2][1]+1) * 128]
random_rgb_img[128:256, 128:256] = rgb_img[order_original[3][0]*128:(order_original[3][0]+1)*128, order_original[3][1]*128:(order_original[3][1]+1)*128]
result = np.concatenate((rgb_img, random_rgb_img), 1)
cv2.imwrite(os.path.join(save, f), result)
print('processing img: %d, name:%s, order:%s' % (n, f, order_original))
n += 1
def random_input9(root, save):
file_list = os.listdir(root)
if not os.path.isdir(save):
os.makedirs(save)
n = 1
for f in tqdm(file_list):
order_original = [[i, j] for i in range(3) for j in range(3)]
rgb_img = cv2.imread(os.path.join(root, f))
print('name:%s' % f)
rgb_img = cv2.resize(rgb_img, (300, 300))
random.shuffle(order_original)
random_rgb_img = np.zeros([300, 300, 3], np.uint8)
num = 0
for i in range(3):
for j in range(3):
random_rgb_img[i*100:(i+1)*100, j*100:(j+1)*100] = rgb_img[order_original[num][0]*100:(order_original[num][0]+1)*100,order_original[num][1]*100:(order_original[num][1]+1)*100]
num += 1
cv2.imwrite(os.path.join(save, f), random_rgb_img)
result = np.concatenate((random_rgb_img, rgb_img), 1)
cv2.imwrite(os.path.join(save, f), result)
print('processing img: %d, name:%s, order:%s' % (n, f, order_original))
n += 1
def random_input9_combine(root, save):
file_list = os.listdir(root)
if not os.path.isdir(save):
os.makedirs(save)
import pdb
pdb.set_trace()
n = 1
for f in tqdm(file_list):
order_original = [[i, j] for i in range(3) for j in range(3)]
rgb_img = cv2.imread(os.path.join(root, f))
print('name:%s' % f)
rgb_img = cv2.resize(rgb_img, (300, 300))
random.shuffle(order_original)
random_rgb_img = np.zeros([300, 300, 3], np.uint8)
num = 0
for i in range(3):
for j in range(3):
random_rgb_img[i*100:(i+1)*100, j*100:(j+1)*100] = rgb_img[order_original[num][0]*100:(order_original[num][0]+1)*100,order_original[num][1]*100:(order_original[num][1]+1)*100]
num += 1
result = np.concatenate((random_rgb_img, rgb_img), 1)
cv2.imwrite(os.path.join(save, f), result)
f_txt = open(save + '/' + '%s.txt' % f, mode='w')
f_txt.write('%s' % order_original[0])
f_txt.close()
print('processing img: %d, name:%s, order:%s' % (n, f, order_original))
n += 1
def edge_promoting(root, save):
file_list = os.listdir(root)
if not os.path.isdir(save):
os.makedirs(save)
n = 1
for f in tqdm(file_list):
rgb_img = cv2.imread(os.path.join(root, f))
rgb_img = cv2.resize(rgb_img, (256, 256))
gauss_img = cv2.GaussianBlur(rgb_img, (5, 5), 2.5)
result = np.concatenate((rgb_img, gauss_img), 1)
cv2.imwrite(os.path.join(save, str(n) + '.png'), result)
#print('processing img: %d' %n)
n += 1