File size: 7,543 Bytes
2252f3d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
import torch.nn as nn
import torch.nn.functional as F


def conv3x3(in_planes, out_planes, stride=1, groups=1, dilation=1):
    """3x3 convolution with padding"""
    return nn.Conv2d(in_planes,
                     out_planes,
                     kernel_size=3,
                     stride=stride,
                     padding=dilation,
                     groups=groups,
                     bias=False,
                     dilation=dilation)


class BasicBlock(nn.Module):
    expansion = 1

    def __init__(self,
                 inplanes,
                 planes,
                 stride=1,
                 downsample=None,
                 groups=1,
                 base_width=64,
                 dilation=1,
                 norm_layer=None,
                 dcn=None):
        super(BasicBlock, self).__init__()
        if norm_layer is None:
            norm_layer = nn.BatchNorm2d
        if groups != 1 or base_width != 64:
            raise ValueError(
                'BasicBlock only supports groups=1 and base_width=64')
        if dilation > 1:
            raise NotImplementedError(
                "Dilation > 1 not supported in BasicBlock")
        # Both self.conv1 and self.downsample layers downsample the input when stride != 1
        self.conv1 = conv3x3(inplanes, planes, stride)
        self.bn1 = norm_layer(planes)
        self.relu = nn.ReLU(inplace=True)
        self.conv2 = conv3x3(planes, planes)
        self.bn2 = norm_layer(planes)
        self.downsample = downsample
        self.stride = stride

    def forward(self, x):
        identity = 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:
            identity = self.downsample(x)

        out += identity
        out = self.relu(out)

        return out


class Bottleneck(nn.Module):
    expansion = 4

    def __init__(self,
                 inplanes,
                 planes,
                 stride=1,
                 downsample=None,
                 norm_layer=nn.BatchNorm2d,
                 dcn=None):
        super(Bottleneck, self).__init__()
        self.dcn = dcn
        self.with_dcn = dcn is not None

        self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=1, bias=False)
        self.bn1 = norm_layer(planes, momentum=0.1)
        self.conv2 = nn.Conv2d(planes,
                               planes,
                               kernel_size=3,
                               stride=stride,
                               padding=1,
                               bias=False)

        self.bn2 = norm_layer(planes, momentum=0.1)
        self.conv3 = nn.Conv2d(planes, planes * 4, kernel_size=1, bias=False)
        self.bn3 = norm_layer(planes * 4, momentum=0.1)
        self.downsample = downsample
        self.stride = stride

    def forward(self, x):
        residual = x

        out = F.relu(self.bn1(self.conv1(x)), inplace=True)
        if not self.with_dcn:
            out = F.relu(self.bn2(self.conv2(out)), inplace=True)
        elif self.with_modulated_dcn:
            offset_mask = self.conv2_offset(out)
            offset = offset_mask[:, :18 * self.deformable_groups, :, :]
            mask = offset_mask[:, -9 * self.deformable_groups:, :, :]
            mask = mask.sigmoid()
            out = F.relu(self.bn2(self.conv2(out, offset, mask)))
        else:
            offset = self.conv2_offset(out)
            out = F.relu(self.bn2(self.conv2(out, offset)), inplace=True)

        out = self.conv3(out)
        out = self.bn3(out)

        if self.downsample is not None:
            residual = self.downsample(x)

        out += residual
        out = F.relu(out)

        return out


class ResNet(nn.Module):
    """ ResNet """

    def __init__(self,
                 architecture,
                 norm_layer=nn.BatchNorm2d,
                 dcn=None,
                 stage_with_dcn=(False, False, False, False)):
        super(ResNet, self).__init__()
        self._norm_layer = norm_layer
        assert architecture in [
            "resnet18", "resnet34", "resnet50", "resnet101", 'resnet152'
        ]
        layers = {
            'resnet18': [2, 2, 2, 2],
            'resnet34': [3, 4, 6, 3],
            'resnet50': [3, 4, 6, 3],
            'resnet101': [3, 4, 23, 3],
            'resnet152': [3, 8, 36, 3],
        }
        self.inplanes = 64
        if architecture == "resnet18" or architecture == 'resnet34':
            self.block = BasicBlock
        else:
            self.block = Bottleneck
        self.layers = layers[architecture]

        self.conv1 = nn.Conv2d(3,
                               64,
                               kernel_size=7,
                               stride=2,
                               padding=3,
                               bias=False)
        self.bn1 = norm_layer(64, eps=1e-5, momentum=0.1, affine=True)
        self.relu = nn.ReLU(inplace=True)
        self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)

        stage_dcn = [dcn if with_dcn else None for with_dcn in stage_with_dcn]

        self.layer1 = self.make_layer(self.block,
                                      64,
                                      self.layers[0],
                                      dcn=stage_dcn[0])
        self.layer2 = self.make_layer(self.block,
                                      128,
                                      self.layers[1],
                                      stride=2,
                                      dcn=stage_dcn[1])
        self.layer3 = self.make_layer(self.block,
                                      256,
                                      self.layers[2],
                                      stride=2,
                                      dcn=stage_dcn[2])

        self.layer4 = self.make_layer(self.block,
                                      512,
                                      self.layers[3],
                                      stride=2,
                                      dcn=stage_dcn[3])

    def forward(self, x):
        x = self.maxpool(self.relu(self.bn1(self.conv1(x))))  # 64 * h/4 * w/4
        x = self.layer1(x)  # 256 * h/4 * w/4
        x = self.layer2(x)  # 512 * h/8 * w/8
        x = self.layer3(x)  # 1024 * h/16 * w/16
        x = self.layer4(x)  # 2048 * h/32 * w/32
        return x

    def stages(self):
        return [self.layer1, self.layer2, self.layer3, self.layer4]

    def make_layer(self, block, planes, blocks, stride=1, dcn=None):
        downsample = None
        if stride != 1 or self.inplanes != planes * block.expansion:
            downsample = nn.Sequential(
                nn.Conv2d(self.inplanes,
                          planes * block.expansion,
                          kernel_size=1,
                          stride=stride,
                          bias=False),
                self._norm_layer(planes * block.expansion),
            )

        layers = []
        layers.append(
            block(self.inplanes,
                  planes,
                  stride,
                  downsample,
                  norm_layer=self._norm_layer,
                  dcn=dcn))
        self.inplanes = planes * block.expansion
        for i in range(1, blocks):
            layers.append(
                block(self.inplanes,
                      planes,
                      norm_layer=self._norm_layer,
                      dcn=dcn))

        return nn.Sequential(*layers)