-
Notifications
You must be signed in to change notification settings - Fork 0
/
sinet.py
240 lines (191 loc) · 9.29 KB
/
sinet.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
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
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
# -*- coding: utf-8 -*-
"""SINet.ipynb
Automatically generated by Colaboratory.
Original file is located at
https://colab.research.google.com/drive/11dSFfMbSu_C71MP91LZONb5gu6QQmBL7
"""
#@title SINET
import torch
import torch.nn as nn
import torchvision.models as models
from .SearchAttention import SA
from Src.backbone.ResNet import ResNet_2Branch
class BasicConv2d(nn.Module):
def __init__(self, in_planes, out_planes, kernel_size, stride=1, padding=0, dilation=1):
super(BasicConv2d, self).__init__()
self.conv = nn.Conv2d(in_planes, out_planes,
kernel_size=kernel_size, stride=stride,
padding=padding, dilation=dilation, bias=False)
self.bn = nn.BatchNorm2d(out_planes)
self.relu = nn.ReLU(inplace=True)
def forward(self, x):
x = self.conv(x)
x = self.bn(x)
return x
class RF(nn.Module):
# Revised from: Receptive Field Block Net for Accurate and Fast Object Detection, 2018, ECCV
# GitHub: https://github.com/ruinmessi/RFBNet
def __init__(self, in_channel, out_channel):
super(RF, self).__init__()
self.relu = nn.ReLU(True)
self.branch0 = nn.Sequential(
BasicConv2d(in_channel, out_channel, 1),
)
self.branch1 = nn.Sequential(
BasicConv2d(in_channel, out_channel, 1),
BasicConv2d(out_channel, out_channel, kernel_size=(1, 3), padding=(0, 1)),
BasicConv2d(out_channel, out_channel, kernel_size=(3, 1), padding=(1, 0)),
BasicConv2d(out_channel, out_channel, 3, padding=3, dilation=3)
)
self.branch2 = nn.Sequential(
BasicConv2d(in_channel, out_channel, 1),
BasicConv2d(out_channel, out_channel, kernel_size=(1, 5), padding=(0, 2)),
BasicConv2d(out_channel, out_channel, kernel_size=(5, 1), padding=(2, 0)),
BasicConv2d(out_channel, out_channel, 3, padding=5, dilation=5)
)
self.branch3 = nn.Sequential(
BasicConv2d(in_channel, out_channel, 1),
BasicConv2d(out_channel, out_channel, kernel_size=(1, 7), padding=(0, 3)),
BasicConv2d(out_channel, out_channel, kernel_size=(7, 1), padding=(3, 0)),
BasicConv2d(out_channel, out_channel, 3, padding=7, dilation=7)
)
self.conv_cat = BasicConv2d(4*out_channel, out_channel, 3, padding=1)
self.conv_res = BasicConv2d(in_channel, out_channel, 1)
def forward(self, x):
x0 = self.branch0(x)
x1 = self.branch1(x)
x2 = self.branch2(x)
x3 = self.branch3(x)
x_cat = self.conv_cat(torch.cat((x0, x1, x2, x3), dim=1))
x = self.relu(x_cat + self.conv_res(x))
return x
class PDC_SM(nn.Module):
# Partial Decoder Component (Search Module)
def __init__(self, channel):
super(PDC_SM, self).__init__()
self.relu = nn.ReLU(True)
self.upsample = nn.Upsample(scale_factor=2, mode='bilinear', align_corners=True)
self.conv_upsample1 = BasicConv2d(channel, channel, 3, padding=1)
self.conv_upsample2 = BasicConv2d(channel, channel, 3, padding=1)
self.conv_upsample3 = BasicConv2d(channel, channel, 3, padding=1)
self.conv_upsample4 = BasicConv2d(channel, channel, 3, padding=1)
self.conv_upsample5 = BasicConv2d(2*channel, 2*channel, 3, padding=1)
self.conv_concat2 = BasicConv2d(2*channel, 2*channel, 3, padding=1)
self.conv_concat3 = BasicConv2d(4*channel, 4*channel, 3, padding=1)
self.conv4 = BasicConv2d(4*channel, 4*channel, 3, padding=1)
self.conv5 = nn.Conv2d(4*channel, 1, 1)
def forward(self, x1, x2, x3, x4):
# print x1.shape, x2.shape, x3.shape, x4.shape
x1_1 = x1
x2_1 = self.conv_upsample1(self.upsample(x1)) * x2
x3_1 = self.conv_upsample2(self.upsample(self.upsample(x1))) * self.conv_upsample3(self.upsample(x2)) * x3
x2_2 = torch.cat((x2_1, self.conv_upsample4(self.upsample(x1_1))), 1)
x2_2 = self.conv_concat2(x2_2)
x3_2 = torch.cat((x3_1, self.conv_upsample5(self.upsample(x2_2)), x4), 1)
x3_2 = self.conv_concat3(x3_2)
x = self.conv4(x3_2)
x = self.conv5(x)
return x
class PDC_IM(nn.Module):
# Partial Decoder Component (Identification Module)
def __init__(self, channel):
super(PDC_IM, self).__init__()
self.relu = nn.ReLU(True)
self.upsample = nn.Upsample(scale_factor=2, mode='bilinear', align_corners=True)
self.conv_upsample1 = BasicConv2d(channel, channel, 3, padding=1)
self.conv_upsample2 = BasicConv2d(channel, channel, 3, padding=1)
self.conv_upsample3 = BasicConv2d(channel, channel, 3, padding=1)
self.conv_upsample4 = BasicConv2d(channel, channel, 3, padding=1)
self.conv_upsample5 = BasicConv2d(2*channel, 2*channel, 3, padding=1)
self.conv_concat2 = BasicConv2d(2*channel, 2*channel, 3, padding=1)
self.conv_concat3 = BasicConv2d(3*channel, 3*channel, 3, padding=1)
self.conv4 = BasicConv2d(3*channel, 3*channel, 3, padding=1)
self.conv5 = nn.Conv2d(3*channel, 1, 1)
def forward(self, x1, x2, x3):
x1_1 = x1
x2_1 = self.conv_upsample1(self.upsample(x1)) * x2
x3_1 = self.conv_upsample2(self.upsample(self.upsample(x1))) * self.conv_upsample3(self.upsample(x2)) * x3
x2_2 = torch.cat((x2_1, self.conv_upsample4(self.upsample(x1_1))), 1)
x2_2 = self.conv_concat2(x2_2)
x3_2 = torch.cat((x3_1, self.conv_upsample5(self.upsample(x2_2))), 1)
x3_2 = self.conv_concat3(x3_2)
x = self.conv4(x3_2)
x = self.conv5(x)
return x
class SINet_ResNet50(nn.Module):
# resnet based encoder decoder
def __init__(self, channel=32, opt=None):
super(SINet_ResNet50, self).__init__()
self.resnet = ResNet_2Branch()
self.downSample = nn.MaxPool2d(2, stride=2)
self.rf_low_sm = RF(320, channel)
self.rf2_sm = RF(3584, channel)
self.rf3_sm = RF(3072, channel)
self.rf4_sm = RF(2048, channel)
self.pdc_sm = PDC_SM(channel)
self.rf2_im = RF(512, channel)
self.rf3_im = RF(1024, channel)
self.rf4_im = RF(2048, channel)
self.pdc_im = PDC_IM(channel)
self.upsample_2 = nn.Upsample(scale_factor=2, mode='bilinear', align_corners=True)
self.upsample_8 = nn.Upsample(scale_factor=8, mode='bilinear', align_corners=True)
self.SA = SA()
if self.training:
self.initialize_weights()
def forward(self, x):
# ---- feature abstraction -----
# - head
x0 = self.resnet.conv1(x)
x0 = self.resnet.bn1(x0)
x0 = self.resnet.relu(x0)
# - low-level features
x0 = self.resnet.maxpool(x0) # (BS, 64, 88, 88)
x1 = self.resnet.layer1(x0) # (BS, 256, 88, 88)
x2 = self.resnet.layer2(x1) # (BS, 512, 44, 44)
# ---- Stage-1: Search Module (SM) ----
x01 = torch.cat((x0, x1), dim=1) # (BS, 64+256, 88, 88)
x01_down = self.downSample(x01) # (BS, 320, 44, 44)
x01_sm_rf = self.rf_low_sm(x01_down) # (BS, 32, 44, 44)
x2_sm = x2 # (512, 44, 44)
x3_sm = self.resnet.layer3_1(x2_sm) # (1024, 22, 22)
x4_sm = self.resnet.layer4_1(x3_sm) # (2048, 11, 11)
x2_sm_cat = torch.cat((x2_sm,
self.upsample_2(x3_sm),
self.upsample_2(self.upsample_2(x4_sm))), dim=1) # 3584 channels
x3_sm_cat = torch.cat((x3_sm,
self.upsample_2(x4_sm)), dim=1) # 3072 channels
x2_sm_rf = self.rf2_sm(x2_sm_cat)
x3_sm_rf = self.rf3_sm(x3_sm_cat)
x4_sm_rf = self.rf4_sm(x4_sm)
camouflage_map_sm = self.pdc_sm(x4_sm_rf, x3_sm_rf, x2_sm_rf, x01_sm_rf)
# ---- Switcher: Search Attention (SA) ----
x2_sa = self.SA(camouflage_map_sm.sigmoid(), x2) # (512, 44, 44)
# ---- Stage-2: Identification Module (IM) ----
x3_im = self.resnet.layer3_2(x2_sa) # (1024, 22, 22)
x4_im = self.resnet.layer4_2(x3_im) # (2048, 11, 11)
x2_im_rf = self.rf2_im(x2_sa)
x3_im_rf = self.rf3_im(x3_im)
x4_im_rf = self.rf4_im(x4_im)
# - decoder
camouflage_map_im = self.pdc_im(x4_im_rf, x3_im_rf, x2_im_rf)
# ---- output ----
return self.upsample_8(camouflage_map_sm), self.upsample_8(camouflage_map_im)
def initialize_weights(self):
resnet50 = models.resnet50(pretrained=True)
pretrained_dict = resnet50.state_dict()
all_params = {}
for k, v in self.resnet.state_dict().items():
if k in pretrained_dict.keys():
v = pretrained_dict[k]
all_params[k] = v
elif '_1' in k:
name = k.split('_1')[0] + k.split('_1')[1]
v = pretrained_dict[name]
all_params[k] = v
elif '_2' in k:
name = k.split('_2')[0] + k.split('_2')[1]
v = pretrained_dict[name]
all_params[k] = v
assert len(all_params.keys()) == len(self.resnet.state_dict().keys())
self.resnet.load_state_dict(all_params)
print('[INFO] initialize weights from resnet50')