-
Notifications
You must be signed in to change notification settings - Fork 0
/
keyselect_pixelshuffle.py
70 lines (62 loc) · 2.52 KB
/
keyselect_pixelshuffle.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
from argparse import ArgumentParser, ArgumentDefaultsHelpFormatter
import numpy as np
import torch
import torchvision
import torchvision.transforms as transforms
import torch.backends.cudnn as cudnn
from utils.pixel_based_encryption import pixel_based_encryption
import lpips
import random
criterion = lpips.LPIPS(net='alex')
device = 'cuda' if torch.cuda.is_available() else 'cpu'
cuda = True if torch.cuda.is_available() else False
if cuda:
criterion = torch.nn.DataParallel(criterion).cuda()
cudnn.benchmark = True
parser = ArgumentParser(formatter_class=ArgumentDefaultsHelpFormatter)
parser.add_argument("--repeat_times", type=int, default=20)
args = parser.parse_args()
# from cifar10 import CIFAR10
trainset = torchvision.datasets.CIFAR10(root='./data_cifar10', train=True, download=True, transform=transforms.Compose([transforms.ToTensor(),]))
trainloader = torch.utils.data.DataLoader(trainset, batch_size=256, shuffle=True, num_workers=16)
def test_lpips(nagaposi, channel_shuffle):
lpips_score, total = 0, 0
for batch_idx, (inputs, targets) in enumerate(trainloader):
images = inputs.numpy().copy()
img = pixel_based_encryption(images, nagaposi, channel_shuffle)
# batch-wise evaluation
lpips_score += torch.sum(criterion.forward(img, inputs)).item()
# image-wise evaluation
# for i in range(img.size()[0]):
# lpips_score += criterion.forward(img[i].view(1,3,32,32), inputs[i].view(1,3,32,32)).item()
total += targets.size(0)
return lpips_score / total
if __name__ == '__main__':
args = parser.parse_args()
inv = np.array([ np.random.randint(0, 2) for i in range(3072)])
color = np.array([ np.random.randint(0, 6) for i in range(1024)])
N_scores = []
for rep_times in range(args.repeat_times+1):
rep_times_check = 10
score_max = 0
N = []
for rep in range(5):
score_max = 0
nagaposi_prop = []
channel_shuffle_prop = []
for tmp in range(rep_times_check):
random.shuffle(inv)
random.shuffle(color)
tmp_score = test_lpips(inv, color)
if score_max == 0:
score_max = tmp_score
nagaposi_prop = inv.copy()
channel_shuffle_prop = color.copy()
elif score_max <= tmp_score:
score_max = tmp_score
nagaposi_prop = inv.copy()
channel_shuffle_prop = color.copy()
N.append(score_max)
print(score_max, N)
N_scores.append(N)
print(N_scores)