-
Notifications
You must be signed in to change notification settings - Fork 12
/
Copy pathresnet_dilated.py
135 lines (106 loc) · 5.14 KB
/
resnet_dilated.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
import torch
import torch.nn as nn
import os
import urllib
# -------------------------------------------------------------------------
# This version of Resnet101 was used in PSPnet and DORN (according to https://github.com/hufu6371/DORN)
# Code based on https://github.com/speedinghzl/pytorch-segmentation-toolbox/blob/master/networks/pspnet.py
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 Bottleneck(nn.Module):
expansion = 4
def __init__(self, inplanes, planes, stride=1, dilation=1, downsample=None, fist_dilation=1):
super(Bottleneck, self).__init__()
self.conv1 = nn.Conv2d(inplanes, planes, kernel_size=1, bias=False)
self.bn1 = nn.BatchNorm2d(planes)
self.conv2 = nn.Conv2d(planes, planes, kernel_size=3, stride=stride,
padding=dilation, dilation=dilation, bias=False)
self.bn2 = nn.BatchNorm2d(planes)
self.conv3 = nn.Conv2d(planes, planes * 4, kernel_size=1, bias=False)
self.bn3 = nn.BatchNorm2d(planes * 4)
self.relu = nn.ReLU(inplace=False)
self.relu_inplace = nn.ReLU(inplace=True)
self.downsample = downsample
self.dilation = dilation
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 = out + residual
out = self.relu_inplace(out)
return out
class ResNet(nn.Module):
def __init__(self, block, layers):
super(ResNet, self).__init__()
self.inplanes = 128
self.conv1 = conv3x3(3, 64, stride=2)
self.bn1 = nn.BatchNorm2d(64)
self.relu1 = nn.ReLU(inplace=False)
self.conv2 = conv3x3(64, 64)
self.bn2 = nn.BatchNorm2d(64)
self.relu2 = nn.ReLU(inplace=False)
self.conv3 = conv3x3(64, 128)
self.bn3 = nn.BatchNorm2d(128)
self.relu3 = nn.ReLU(inplace=False)
self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1, ceil_mode=True) # change
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)
def _make_layer(self, block, planes, blocks, stride=1, dilation=1):
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),
nn.BatchNorm2d(planes * block.expansion,affine = True))
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 forward(self, x):
x = self.relu1(self.bn1(self.conv1(x)))
x = self.relu2(self.bn2(self.conv2(x)))
x = self.relu3(self.bn3(self.conv3(x)))
x = self.maxpool(x)
x = self.layer1(x)
x = self.layer2(x)
x = self.layer3(x)
x = self.layer4(x)
return x
def resnet101dilated(pretrained=False):
model = ResNet(Bottleneck,[3, 4, 23, 3])
if pretrained:
# Download pretrained model if it does not exist
# This is ADE20K-pretrained encoder from https://github.com/CSAILVision/semantic-segmentation-pytorch
# It is the only pretrained resnet101dilated I could find online
# (with the same architecture as in the paper - see https://github.com/hufu6371/DORN/tree/master/models)
filename = './pretrained/ade20k_resnet101dilated_encoder_epoch_25.pth'
if not os.path.isfile(filename):
os.makedirs(os.path.dirname(filename))
print('Pretrained feature extractor not found. Downloading...')
url = 'http://sceneparsing.csail.mit.edu/model/pytorch/ade20k-resnet101dilated-ppm_deepsup/encoder_epoch_25.pth'
urllib.request.urlretrieve(url, filename)
print('Download completed:', filename)
# Load pretrained parameters
saved_state_dict = torch.load(filename, map_location='cpu')
new_params = model.state_dict().copy()
for i in saved_state_dict:
i_parts = i.split('.')
if not i_parts[0] == 'fc' and i.find('._') == -1:
new_params['.'.join(i_parts[0:])] = saved_state_dict[i]
model.load_state_dict(new_params)
return model