-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathResNet18Block.py
127 lines (108 loc) · 4.18 KB
/
ResNet18Block.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
import torch
import torch.nn as nn
import time
import torch.nn.functional as F
# 块层数，第一层一个，剩余为块内层数
def_layer = [64, 64, 64, 128, 128, 256, 256, 512, 512]
class Feature_enhancement(nn.Module):
def __init__(self, channel):
super(Feature_enhancement, self).__init__()
self.globalAvgpool = nn.AdaptiveAvgPool2d(1)
self.fc = nn.Linear(channel, channel, bias=False)
self.sigmoid = nn.Sigmoid()
def forward(self, x):
out = self.globalAvgpool(x)
out = out.view(out.size(0), -1)
out = self.fc(out)
out = self.sigmoid(out)
out = out.view(out.size(0), out.size(1), 1, 1)
out = out*x
return out
class BasicBlock(nn.Module):
def __init__(self, inchannel, outchannel, strid):
super(BasicBlock, self).__init__()
self.strid_ = strid
temp_outchannel = int(outchannel/8)
self.conv1 = nn.Sequential(
nn.Conv2d(inchannel, temp_outchannel, kernel_size=1, padding=0, bias=False),
nn.Conv2d(temp_outchannel, temp_outchannel, kernel_size=3, stride=strid, padding=0, bias=False),
Feature_enhancement(temp_outchannel),
nn.Conv2d(temp_outchannel, outchannel, kernel_size=1, padding=0, bias=False),
)
self.conv2 = nn.Sequential(
nn.Conv2d(outchannel, temp_outchannel, kernel_size=1, padding=0, bias=False),
nn.Conv2d(temp_outchannel, temp_outchannel, kernel_size=3, stride=1, padding=0, bias=False),
Feature_enhancement(temp_outchannel),
nn.Conv2d(temp_outchannel, outchannel, kernel_size=1, padding=0, bias=False),
nn.BatchNorm2d(outchannel),
nn.ReLU(inplace=True)
)
self.shortcut = nn.Sequential(
nn.Conv2d(inchannel, outchannel, kernel_size=1, stride=strid, padding=0, bias=False),
nn.BatchNorm2d(outchannel)
)
def forward(self, x):
out = self.conv1(x)
out = self.conv2(out)
if self.strid_ == 2:
xsout = x[:, :, 3:x.shape[2] - 3, 3:x.shape[3] - 3]
else:
xsout = x[:, :, 2:x.shape[2] - 2, 2:x.shape[3] - 2]
sout = self.shortcut(xsout)
out = out + sout
# out = x
# if self.strid_ != 2:
# out = self.conv1(x)
# out = self.conv2(out)
# if self.strid_ == 2:
# xsout = x[:, :, 1:x.shape[2] - 1, 1:x.shape[3] - 1]
# else:
# xsout = x[:, :, 2:x.shape[2] - 2, 2:x.shape[3] - 2]
# if self.strid_ != 2:
# sout = self.shortcut(xsout)
# else:
# sout = xsout
# out = out + sout
out = F.relu(out)
return out
class resnet18Blockx(nn.Module):
def __init__(self, layer_num=None):
super(resnet18Blockx, self).__init__()
if layer_num == None:
self.layer_num = def_layer
else:
self.layer_num = layer_num
self.feature = self._make_layer(BasicBlock)
self.fc = nn.Linear(self.layer_num[-1] * 3 * 3, 6)
def forward(self, x):
index = 0
for module in self.feature:
if 5 <= index < 9:
x = module(x)
index += 1
# out = F.avg_pool2d(x, 2)
# out = out.view(out.size(0), -1)
# out = self.fc(out)
return x
def _make_layer(self, BasicBlock):
self.inchannel = 3
strid = [2, 1, 2, 1, 2, 1, 2, 1]
layer = []
for i in range(len(self.layer_num)):
outchannel = self.layer_num[i]
if i == 0:
temp_layer = nn.Sequential(
nn.Conv2d(self.inchannel, outchannel, kernel_size=3, stride=2, padding=0, bias=False),
nn.BatchNorm2d(self.layer_num[i]),
nn.ReLU(inplace=True)
)
else:
temp_layer = BasicBlock(self.inchannel, outchannel, strid[i - 1])
layer.append(temp_layer)
self.inchannel = outchannel
return nn.Sequential(*layer)
if __name__ == "__main__":
layer_name = [64, 64, 64, 128, 128, 256, 256, 512, 512]
net = resnet18(layer_name)
x = torch.rand([1, 3, 200, 200])
print(net)