-
Notifications
You must be signed in to change notification settings - Fork 4
Expand file tree
/
Copy pathclassify.py
More file actions
200 lines (167 loc) · 6.13 KB
/
classify.py
File metadata and controls
200 lines (167 loc) · 6.13 KB
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
import torch.nn as nn
import torchvision
import evolve
############################# Normal #############################
class Flatten(nn.Module):
def forward(self, input):
return input.view(input.size(0), -1)
class VGG16(nn.Module):
def __init__(self, n_classes):
super(VGG16, self).__init__()
model = torchvision.models.vgg16_bn(pretrained=True)
self.feature = model.features
self.feat_dim = 512 * 2 * 2
self.n_classes = n_classes
self.bn = nn.BatchNorm1d(self.feat_dim)
self.bn.bias.requires_grad_(False) # no shift
self.fc_layer = nn.Linear(self.feat_dim, self.n_classes)
def forward(self, x):
feature = self.feature(x)
feature = feature.view(feature.size(0), -1)
feature = self.bn(feature)
res = self.fc_layer(feature)
return feature, res
class FaceNet64(nn.Module):
def __init__(self, num_classes = 1000):
super(FaceNet64, self).__init__()
self.feature = evolve.IR_50_64((64, 64))
self.feat_dim = 512
self.num_classes = num_classes
self.output_layer = nn.Sequential(nn.BatchNorm2d(512),
nn.Dropout(),
Flatten(),
nn.Linear(512 * 4 * 4, 512),
nn.BatchNorm1d(512))
self.fc_layer = nn.Linear(self.feat_dim, self.num_classes)
def forward(self, x):
feat = self.feature(x)
feat = self.output_layer(feat)
feat = feat.view(feat.size(0), -1)
out = self.fc_layer(feat)
return feat, out
class IR152(nn.Module):
def __init__(self, num_classes=1000):
super(IR152, self).__init__()
self.feature = evolve.IR_152_64((64, 64))
self.feat_dim = 512
self.num_classes = num_classes
self.output_layer = nn.Sequential(nn.BatchNorm2d(512),
nn.Dropout(),
Flatten(),
nn.Linear(512 * 4 * 4, 512),
nn.BatchNorm1d(512))
self.fc_layer = nn.Linear(self.feat_dim, self.num_classes)
def forward(self, x):
feat = self.feature(x)
feat = self.output_layer(feat)
feat = feat.view(feat.size(0), -1)
out = self.fc_layer(feat)
return feat, out
class IR18(nn.Module):
def __init__(self, num_classes=5):
super(IR18, self).__init__()
model = torchvision.models.resnet18(pretrained=True)
self.feature = nn.Sequential(
model.conv1,
model.bn1,
model.relu,
model.maxpool,
model.layer1,
model.layer2,
model.layer3,
model.layer4,
model.avgpool,
Flatten()
)
self.fc_layer = nn.Linear(model.fc.in_features, num_classes)
def forward(self, x):
feat = self.feature(x)
out = self.fc_layer(feat)
return feat, out
############################# BiDO #############################
cfgs = {
'A': [64, 'M', 128, 'M', 256, 256, 'M', 512, 512, 'M', 512, 512, 'M'],
'B': [64, 64, 'M', 128, 128, 'M', 256, 256, 'M', 512, 512, 'M', 512, 512, 'M'],
'D': [64, 64, 'M', 128, 128, 'M', 256, 256, 256, 'M', 512, 512, 512, 'M', 512, 512, 512, 'M'],
'E': [64, 64, 'M', 128, 128, 'M', 256, 256, 256, 256, 'M', 512, 512, 512, 512, 'M', 512, 512, 512, 512, 'M'],
}
def make_layers(cfg, batch_norm=False):
blocks = []
layers = []
in_channels = 3
for v in cfg:
if v == 'M':
layers += [nn.MaxPool2d(kernel_size=2, stride=2)]
blocks.append(nn.Sequential(*layers))
layers = []
else:
conv2d = nn.Conv2d(in_channels, v, kernel_size=3, padding=1)
if batch_norm:
layers += [conv2d, nn.BatchNorm2d(v), nn.ReLU(inplace=True)]
else:
layers += [conv2d, nn.ReLU(inplace=True)]
in_channels = v
return blocks
class VGG16_BiDO(nn.Module):
def __init__(self, n_classes):
super(VGG16_BiDO, self).__init__()
blocks = make_layers(cfgs['D'], batch_norm=True)
self.layer1 = blocks[0]
self.layer2 = blocks[1]
self.layer3 = blocks[2]
self.layer4 = blocks[3]
self.layer5 = blocks[4]
self.feat_dim = 512 * 2 * 2
self.n_classes = n_classes
self.bn = nn.BatchNorm1d(self.feat_dim)
self.bn.bias.requires_grad_(False) # no shift
self.fc_layer = nn.Linear(self.feat_dim, self.n_classes)
def forward(self, x):
hiddens = []
out = self.layer1(x)
hiddens.append(out)
out = self.layer2(out)
hiddens.append(out)
out = self.layer3(out)
hiddens.append(out)
out = self.layer4(out)
hiddens.append(out)
feature = self.layer5(out)
feature = feature.view(feature.size(0), -1)
feature = self.bn(feature)
hiddens.append(feature)
res = self.fc_layer(feature)
return hiddens, res
class IR18_BiDO(nn.Module):
def __init__(self, num_classes=5):
super(IR18_BiDO, self).__init__()
model = torchvision.models.resnet18(pretrained=True)
self.input_layer = nn.Sequential(
model.conv1,
model.bn1,
model.relu,
model.maxpool
)
self.layer1 = model.layer1
self.layer2 = model.layer2
self.layer3 = model.layer3
self.layer4 = model.layer4
self.output_layer = nn.Sequential(
model.avgpool,
Flatten()
)
self.fc_layer = nn.Linear(model.fc.in_features, num_classes)
def forward(self, x):
hiddens = []
feat = self.input_layer(x)
feat = self.layer1(feat)
hiddens.append(feat)
feat = self.layer2(feat)
hiddens.append(feat)
feat = self.layer3(feat)
hiddens.append(feat)
feat = self.layer4(feat)
hiddens.append(feat)
feat = self.output_layer(feat)
out = self.fc_layer(feat)
return hiddens, out