-
Notifications
You must be signed in to change notification settings - Fork 43
/
m3d.py
177 lines (141 loc) · 6.82 KB
/
m3d.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
import math
import torch
import torch.nn as nn
from torch import Tensor
from ..gradient.mifgsm import MIFGSM
from ..utils import *
class M3D(MIFGSM):
"""
M3D Attack
'Minimizing Maximum Model Discrepancy for Transferable Black-box Targeted Attacks (CVPR 2023)'(https://openaccess.thecvf.com/content/CVPR2023/papers/Zhao_Minimizing_Maximum_Model_Discrepancy_for_Transferable_Black-Box_Targeted_Attacks_CVPR_2023_paper.pdf)
Arguments:
model_name (str): the surrogate model name for attack.
Example script:
python main.py --input_dir ./path/to/data --output_dir adv_data/m3d/resnet50 --attack m3d --model=resnet50
python main.py --input_dir ./path/to/data --output_dir adv_data/m3d/resnet50 --attack m3d --model=resnet50 --eval
"""
def __init__(self, model_name="resnet18", *args, **kwargs):
super().__init__(model_name, *args, **kwargs)
self.model_name = model_name
self.netG_list = []
for target_class in generation_target_classes:
self.netG_list.append(self.load_Gmodel(target_class))
def load_Gmodel(self, target_class):
netG = GeneratorResnet()
file_path = "/path/to/checkpoint/m3d/netG_{}_9_{}.pth".format(self.model_name, target_class)
try:
netG.load_state_dict(torch.load(file_path))
except:
raise FileExistsError(
f"No pre-trained generator model found at {file_path}, please visit "
"https://github.com/Asteriajojo/M3D or "
"https://huggingface.co/Trustworthy-AI-Group/TransferAttack/blob/main/M3D.zip "
"to download the model."
)
netG.to(self.device)
netG.eval()
return netG
def forward(self, data: Tensor, label: Tensor, idx, **kwargs):
netG = self.netG_list[idx]
data = data.clone().detach().to(self.device)
kernel_size = 3
pad = 2
sigma = 1
kernel = get_gaussian_kernel(kernel_size=kernel_size, pad=pad, sigma=sigma).cuda()
with torch.no_grad():
adv_imgs = netG(data).detach()
adv_imgs = kernel(adv_imgs)
perturbations = adv_imgs - data
perturbations = torch.clamp(perturbations, -self.epsilon, self.epsilon)
return perturbations
###########################
# Generator: Resnet
###########################
ngf = 64
class GeneratorResnet(nn.Module):
def __init__(self, inception=False):
"""
:param inception: if True crop layer will be added to go from 3x300x300 t0 3x299x299.
:param data_dim: for high dimentional dataset (imagenet) 6 resblocks will be add otherwise only 2.
"""
super(GeneratorResnet, self).__init__()
self.inception = inception
# Input_size = 3, n, n
self.block1 = nn.Sequential(nn.ReflectionPad2d(3), nn.Conv2d(3, ngf, kernel_size=7, padding=0, bias=False), nn.BatchNorm2d(ngf), nn.ReLU(True))
# Input size = 3, n, n
self.block2 = nn.Sequential(nn.Conv2d(ngf, ngf * 2, kernel_size=3, stride=2, padding=1, bias=False), nn.BatchNorm2d(ngf * 2), nn.ReLU(True))
# Input size = 3, n/2, n/2
self.block3 = nn.Sequential(nn.Conv2d(ngf * 2, ngf * 4, kernel_size=3, stride=2, padding=1, bias=False), nn.BatchNorm2d(ngf * 4), nn.ReLU(True))
# Input size = 3, n/4, n/4
# Residual Blocks: 6
self.resblock1 = ResidualBlock(ngf * 4)
self.resblock2 = ResidualBlock(ngf * 4)
self.resblock3 = ResidualBlock(ngf * 4)
self.resblock4 = ResidualBlock(ngf * 4)
self.resblock5 = ResidualBlock(ngf * 4)
self.resblock6 = ResidualBlock(ngf * 4)
# Input size = 3, n/4, n/4
self.upsampl1 = nn.Sequential(
nn.ConvTranspose2d(ngf * 4, ngf * 2, kernel_size=3, stride=2, padding=1, output_padding=1, bias=False), nn.BatchNorm2d(ngf * 2), nn.ReLU(True)
)
# Input size = 3, n/2, n/2
self.upsampl2 = nn.Sequential(
nn.ConvTranspose2d(ngf * 2, ngf, kernel_size=3, stride=2, padding=1, output_padding=1, bias=False), nn.BatchNorm2d(ngf), nn.ReLU(True)
)
# Input size = 3, n, n
self.blockf = nn.Sequential(nn.ReflectionPad2d(3), nn.Conv2d(ngf, 3, kernel_size=7, padding=0))
self.crop = nn.ConstantPad2d((0, -1, -1, 0), 0)
def forward(self, input):
x = self.block1(input)
x = self.block2(x)
x = self.block3(x)
x = self.resblock1(x)
x = self.resblock2(x)
x = self.resblock3(x)
x = self.resblock4(x)
x = self.resblock5(x)
x = self.resblock6(x)
x = self.upsampl1(x)
x = self.upsampl2(x)
x = self.blockf(x)
if self.inception:
x = self.crop(x)
return (torch.tanh(x) + 1) / 2 # Output range [0 1]
class ResidualBlock(nn.Module):
def __init__(self, num_filters):
super(ResidualBlock, self).__init__()
self.block = nn.Sequential(
nn.ReflectionPad2d(1),
nn.Conv2d(in_channels=num_filters, out_channels=num_filters, kernel_size=3, stride=1, padding=0, bias=False),
nn.BatchNorm2d(num_filters),
nn.ReLU(True),
nn.Dropout(0.5),
nn.ReflectionPad2d(1),
nn.Conv2d(in_channels=num_filters, out_channels=num_filters, kernel_size=3, stride=1, padding=0, bias=False),
nn.BatchNorm2d(num_filters),
)
def forward(self, x):
residual = self.block(x)
return x + residual
# From https://discuss.pytorch.org/t/is-there-anyway-to-do-gaussian-filtering-for-an-image-2d-3d-in-pytorch/12351/3
def get_gaussian_kernel(kernel_size=3, pad=2, sigma=2, channels=3):
# Create a x, y coordinate grid of shape (kernel_size, kernel_size, 2)
x_coord = torch.arange(kernel_size)
x_grid = x_coord.repeat(kernel_size).view(kernel_size, kernel_size)
y_grid = x_grid.t()
xy_grid = torch.stack([x_grid, y_grid], dim=-1).float()
mean = (kernel_size - 1) / 2.0
variance = sigma**2.0
# Calculate the 2-dimensional gaussian kernel which is
# the product of two gaussian distributions for two different
# variables (in this case called x and y)
gaussian_kernel = (1.0 / (2.0 * math.pi * variance)) * torch.exp(-torch.sum((xy_grid - mean) ** 2.0, dim=-1) / (2 * variance))
# Make sure sum of values in gaussian kernel equals 1.
gaussian_kernel = gaussian_kernel / torch.sum(gaussian_kernel)
# Reshape to 2d depthwise convolutional weight
gaussian_kernel = gaussian_kernel.view(1, 1, kernel_size, kernel_size)
gaussian_kernel = gaussian_kernel.repeat(channels, 1, 1, 1)
gaussian_filter = nn.Conv2d(in_channels=channels, out_channels=channels, kernel_size=kernel_size, groups=channels, padding=kernel_size - pad, bias=False)
gaussian_filter.weight.data = gaussian_kernel
gaussian_filter.weight.requires_grad = False
return gaussian_filter