-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathmodel.py
177 lines (132 loc) · 5.76 KB
/
model.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 torch
from torch import nn
from torch.nn import init
import torch.nn.functional as F
def _weights_init(m):
if isinstance(m, nn.Linear) or isinstance(m, nn.Conv2d):
init.kaiming_normal_(m.weight)
class LambdaLayerCifar(nn.Module):
def __init__(self, lambd):
super(LambdaLayerCifar, self).__init__()
self.lambd = lambd
def forward(self, x):
return self.lambd(x)
class BasicBlockCifar(nn.Module):
expansion = 1
def __init__(self, in_planes, planes, stride=1, option='A'):
super(BasicBlockCifar, self).__init__()
self.conv1 = nn.Conv2d(in_planes,
planes,
kernel_size=3,
stride=stride,
padding=1,
bias=False)
self.bn1 = nn.BatchNorm2d(planes)
self.conv2 = nn.Conv2d(planes,
planes,
kernel_size=3,
stride=1,
padding=1,
bias=False)
self.bn2 = nn.BatchNorm2d(planes)
self.shortcut = nn.Sequential()
self.relu = nn.ReLU()
if stride != 1 or in_planes != planes:
if option == 'A':
tup1 = (0, 0, 0, 0, planes // 4, planes // 4)
def f(x): return F.pad(x[:, :, ::2, ::2], tup1, 'constant', 0)
self.shortcut = LambdaLayerCifar(f)
def forward(self, x):
out = self.relu(self.bn1(self.conv1(x)))
out = self.bn2(self.conv2(out))
out += self.shortcut(x)
out = self.relu(out)
return out
class ResNetCifar(nn.Module):
def __init__(self, block, num_blocks, num_classes=10):
super(ResNetCifar, self).__init__()
self.in_planes = 16
self.conv1 = nn.Conv2d(3,
16,
kernel_size=3,
stride=1,
padding=1,
bias=False)
self.bn1 = nn.BatchNorm2d(16)
self.layer1 = self._make_layer(block, 16, num_blocks[0], stride=1)
self.layer2 = self._make_layer(block, 32, num_blocks[1], stride=2)
self.layer3 = self._make_layer(block, 64, num_blocks[2], stride=2)
self.fc = nn.Linear(64, num_classes)
self.relu = nn.ReLU()
self.apply(_weights_init)
def _make_layer(self, block, planes, num_blocks, stride):
strides = [stride] + [1] * (num_blocks - 1)
layers = []
for stride in strides:
layers.append(block(self.in_planes, planes, stride))
self.in_planes = planes * block.expansion
return nn.Sequential(*layers)
def forward(self, x):
out = self.relu(self.bn1(self.conv1(x)))
out = self.layer1(out)
out = self.layer2(out)
out = self.layer3(out)
out = F.avg_pool2d(out, out.size()[3])
out = out.view(out.size(0), -1)
out = self.fc(out)
return out
class ResNet14(nn.Module):
def __init__(self, inc, outc):
super().__init__()
self.left_model = ResNetCifar(BasicBlockCifar, [2, 2, 2])
self.right_model = ResNetCifar(BasicBlockCifar, [2, 2, 2])
self.left_model.conv1 = nn.Conv2d(inc,
16,
kernel_size=3,
stride=1,
padding=1,
bias=False)
self.right_model.conv1 = nn.Conv2d(inc,
16,
kernel_size=3,
stride=1,
padding=1,
bias=False)
self.left_model.fc = nn.Linear(64, 32, bias=True)
self.left_dropout = nn.Dropout(p=0.5)
self.right_model.fc = nn.Linear(64, 32, bias=True)
self.right_dropout = nn.Dropout(p=0.5)
self.fc = nn.Linear(64, outc, bias=True)
def forward(self, left_eye, right_eye):
left_out = self.left_dropout(self.left_model(left_eye))
right_out = self.right_dropout(self.right_model(right_eye))
out = self.fc(torch.cat((left_out, right_out), dim=1))
return out
class ResNet20(nn.Module):
def __init__(self, inc, outc):
super().__init__()
self.left_model = ResNetCifar(BasicBlockCifar, [3, 3, 3])
self.right_model = ResNetCifar(BasicBlockCifar, [3, 3, 3])
self.left_model.conv1 = nn.Conv2d(inc,
16,
kernel_size=3,
stride=1,
padding=1,
bias=False)
self.right_model.conv1 = nn.Conv2d(inc,
16,
kernel_size=3,
stride=1,
padding=1,
bias=False)
self.left_model.fc = nn.Linear(64, 32, bias=True)
self.left_dropout = nn.Dropout()
self.right_model.fc = nn.Linear(64, 32, bias=True)
self.right_dropout = nn.Dropout()
self.fc = nn.Linear(64, outc, bias=True)
self.relu = nn.ReLU()
def forward(self, left_eye, right_eye):
left_out = self.left_dropout(self.left_model(left_eye))
right_out = self.right_dropout(self.right_model(right_eye))
out = self.relu(self.fc(torch.cat((left_out, right_out), dim=1)))
return out