-
Notifications
You must be signed in to change notification settings - Fork 6
/
resnet.py
192 lines (151 loc) · 6.24 KB
/
resnet.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
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
import torch.nn as nn
import torch
import numpy as np
import torchvision
import torch.nn.functional as F
# code of dilated convolution part is referenced from https://github.com/speedinghzl/Pytorch-Deeplab
affine_par = True
def outS(i):
i = int(i)
i = (i + 1) / 2
i = int(np.ceil((i + 1) / 2.0))
i = (i + 1) / 2
return i
def conv3x3(in_planes, out_planes, stride=1):
"3x3 convolution with padding"
return nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride,
padding=1, bias=False)
class BasicBlock(nn.Module):
expansion = 1
def __init__(self, inplanes, planes, stride=1, downsample=None):
super(BasicBlock, self).__init__()
self.conv1 = conv3x3(inplanes, planes, stride)
self.bn1 = nn.BatchNorm2d(planes, affine=affine_par)
self.relu = nn.ReLU(inplace=True)
self.conv2 = conv3x3(planes, planes)
self.bn2 = nn.BatchNorm2d(planes, affine=affine_par)
self.downsample = downsample
self.stride = stride
def forward(self, x):
residual = x
out = self.conv1(x)
out = self.bn1(out)
out = self.relu(out)
out = self.conv2(out)
out = self.bn2(out)
if self.downsample is not None:
residual = self.downsample(x)
out += residual
out = self.relu(out)
return out
class Bottleneck(nn.Module):
expansion = 4
def __init__(self, inplanes, planes, stride=1, dilation=1, downsample=None):
super(Bottleneck, self).__init__()
self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=1, stride=stride, bias=False)
self.bn1 = nn.BatchNorm2d(planes, affine=affine_par)
for i in self.bn1.parameters():
i.requires_grad = False
padding = dilation
self.conv2 = nn.Conv2d(planes, planes, kernel_size=3, stride=1,
padding=padding, bias=False, dilation=dilation)
self.bn2 = nn.BatchNorm2d(planes, affine=affine_par)
for i in self.bn2.parameters():
i.requires_grad = False
self.conv3 = nn.Conv2d(planes, planes * 4, kernel_size=1, bias=False)
self.bn3 = nn.BatchNorm2d(planes * 4, affine=affine_par)
for i in self.bn3.parameters():
i.requires_grad = False
self.relu = nn.ReLU(inplace=True)
self.downsample = downsample
self.stride = stride
def forward(self, x):
residual = x
out = self.conv1(x)
out = self.bn1(out)
out = self.relu(out)
out = self.conv2(out)
out = self.bn2(out)
out = self.relu(out)
out = self.conv3(out)
out = self.bn3(out)
if self.downsample is not None:
residual = self.downsample(x)
out += residual
out = self.relu(out)
return out
class ResNet(nn.Module):
def __init__(self, block, layers, num_classes):
self.inplanes = 64
super(ResNet, self).__init__()
self.conv1 = nn.Conv2d(3, 64, kernel_size=7, stride=2, padding=3,
bias=False)
self.bn1 = nn.BatchNorm2d(64, affine=affine_par)
for i in self.bn1.parameters():
i.requires_grad = False
self.relu = nn.ReLU(inplace=True)
self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1, ceil_mode=True)
self.layer1 = self._make_layer(block, 64, layers[0])
self.layer2 = self._make_layer(block, 128, layers[1], stride=2)
self.layer3 = self._make_layer(block, 256, layers[2], stride=1, dilation=2)
self.layer4 = self._make_layer(block, 512, layers[3], stride=1, dilation=4)
for m in self.modules():
if isinstance(m, nn.Conv2d):
n = m.kernel_size[0] * m.kernel_size[1] * m.out_channels
m.weight.data.normal_(0, 0.01)
elif isinstance(m, nn.BatchNorm2d):
m.weight.data.fill_(1)
m.bias.data.zero_()
def _make_layer(self, block, planes, blocks, stride=1, dilation=1):
downsample = None
if stride != 1 or self.inplanes != planes * block.expansion or dilation == 2 or dilation == 4: # 注意,resnet中每个大block内的第一个bottleneck之后又个downsample
downsample = nn.Sequential(
nn.Conv2d(self.inplanes, planes * block.expansion,
kernel_size=1, stride=stride, bias=False),
nn.BatchNorm2d(planes * block.expansion, affine=affine_par))
for i in downsample._modules['1'].parameters():
i.requires_grad = False
layers = []
layers.append(block(self.inplanes, planes, stride, dilation=dilation, downsample=downsample))
self.inplanes = planes * block.expansion
for i in range(1, blocks):
layers.append(block(self.inplanes, planes, dilation=dilation))
return nn.Sequential(*layers)
def _make_pred_layer(self, block, dilation_series, padding_series, num_classes):
return block(dilation_series, padding_series, num_classes)
def forward(self, x):
# important: do not optimize the RESNET backbone
x = self.conv1(x)
x = self.bn1(x)
x = self.relu(x)
x = self.maxpool(x)
out1 = self.layer1(x)
out2 = self.layer2(out1)
out3 = self.layer3(out2)
out4 = self.layer4(out3)
outputs = []
outputs.append(out1)
outputs.append(out2)
outputs.append(out3)
outputs.append(out4)
# out4 = self.layer4(out3)
# outputs.append(out4)
return outputs
def load_resnet50_param(model, stop_layer='layer4'):
resnet50 = torchvision.models.resnet50(pretrained=True)
saved_state_dict = resnet50.state_dict()
new_params = model.state_dict().copy()
for i in saved_state_dict: # copy params from resnet50,except layers after stop_layer
i_parts = i.split('.')
if not i_parts[0] == stop_layer:
new_params['.'.join(i_parts)] = saved_state_dict[i]
else:
break
model.load_state_dict(new_params)
model.train()
return model
def Res_Deeplab(num_classes=2, pretrained=True):
model = ResNet(Bottleneck, [3, 4, 6, 3], num_classes)
if pretrained:
model = load_resnet50_param(model)
return model