-
Notifications
You must be signed in to change notification settings - Fork 5
/
Copy pathIGCV3.py
167 lines (148 loc) · 6.2 KB
/
IGCV3.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
#**coding=utf-8**
import torch.nn as nn
import torch.functional
import math
#IGCV3: Interleaved Low-Rank Group Convolutions for Efficient Deep Neural Networks. Ke Sun, Mingjie Li, Dong Liu, and Jingdong Wang. arXiv preprint arXIV:1806.00178 (2017)
#
def conv_bn(inp, oup, stride ):
return nn.Sequential(
nn.Conv2d(inp, oup,kernel_size= 3, stride= stride, padding= 1, bias=False),
nn.BatchNorm2d(oup),
nn.ReLU6(inplace=True)
)
def conv_1x1_bn(inp, oup):
return nn.Sequential(
nn.Conv2d(inp, oup, kernel_size = 1, stride= 1, padding= 0, bias=False),
nn.BatchNorm2d(oup),
nn.ReLU6(inplace=True)
)
class PermutationBlock(nn.Module):
def __init__(self, groups):
super(PermutationBlock, self).__init__()
self.groups = groups
def forward(self, input):
n, c, h, w = input.size()
G = self.groups
#直接就是mxnet实现的permutation操作
# def permutation(data, groups):
#举例说明：当groups = 2时，输入：nx144x56x56
# data = mx.sym.reshape(data, shape=(0, -4, groups, -1, -2))
# 输出：nx2x72x56x56
# data = mx.sym.swapaxes(data, 1, 2)
# 输出：nx72x2x56x56
# data = mx.sym.reshape(data, shape=(0, -3, -2))
# 输出：nx144x56x56
# return data
output = input.view(n, G, c // G, h, w).permute(0, 2, 1, 3, 4).contiguous().view(n, c, h, w)
return output
class InvertedResidual(nn.Module):
def __init__(self, inp, oup, stride, expand_ratio):
super(InvertedResidual, self).__init__()
self.stride = stride
assert stride in [1, 2]
self.use_res_connect = self.stride == 1 and inp == oup
self.conv = nn.Sequential(
# pw
nn.Conv2d(inp, inp * expand_ratio,kernel_size = 1, stride= 1, padding=0,groups = 2, bias=False),
nn.BatchNorm2d(inp * expand_ratio),
nn.ReLU6(inplace=True),
#permutation
PermutationBlock(groups=2),
# dw
nn.Conv2d(inp * expand_ratio, inp * expand_ratio, kernel_size =3, stride= stride, padding=1, groups=inp * expand_ratio, bias=False),
nn.BatchNorm2d(inp * expand_ratio),
nn.ReLU6(inplace=True),
# pw-linear
nn.Conv2d(inp * expand_ratio, oup, kernel_size =1, stride= 1, padding=0,groups = 2, bias=False),
nn.BatchNorm2d(oup),
# permutation
PermutationBlock(groups= int(round((oup/2)))),
)
def forward(self, x):
if self.use_res_connect:
return x + self.conv(x)
else:
return self.conv(x)
class IGCV3(nn.Module):
def __init__(self,args):
super(IGCV3, self).__init__()
# 配置某些block的stride，满足downsampling的需求
s1, s2 = 2, 2
if args.downsampling == 16:
s1, s2 = 2, 1
elif args.downsampling == 8:
s1, s2 = 1, 1
'''
network_settings网络的相关配置，从该参数可以看出，Mobile-Net由9个部分组成,
姑且叫做Mobile block。
network_settings中:
't'表示Inverted Residuals的扩征系数
'c'表示该block输出的通道数
‘n’表示当前block由几个残差单元组成
's'表示当前block的stride
'''
# setting of inverted residual blocks
self.interverted_residual_setting = [
# t, c, n, s
[1, 16, 1, 1],
[6, 24, 4, s2],
[6, 32, 6, 2],
[6, 64, 8, 2],
[6, 96, 6, 1],
[6, 160, 6, 2],
[6, 320, 1, 1],
]
# building first layer
assert args.img_height % 32 == 0
input_channel = int(32 * args.width_multiplier)
self.last_channel = int(1280 * args.width_multiplier) if args.width_multiplier > 1.0 else 1280
#第一层，
self.features = [conv_bn(inp =3, oup =input_channel, stride = s1)]
#中间block，一共7个,
# Layers from 1 to 7
# building inverted residual blocks
for t, c, n, s in self.interverted_residual_setting:
output_channel = int(c * args.width_multiplier)
for i in range(n):
if i == 0:
self.features.append(InvertedResidual(input_channel, output_channel, s, t))
else:
self.features.append(InvertedResidual(input_channel, output_channel, 1, t))
input_channel = output_channel
# building last several layers
self.features.append(conv_1x1_bn(input_channel, self.last_channel))
self.features.append(nn.AvgPool2d(kernel_size = (args.img_height // args.downsampling, args.img_width // args.downsampling)))
# make it nn.Sequential
self.features = nn.Sequential(*self.features)
# building classifier
self.classifier = nn.Sequential(
nn.Dropout(),
nn.Linear(self.last_channel,args.num_classes),
)
self._initialize_weights()
def forward(self, x):
x = self.features(x)
x = x.view(-1, self.last_channel)
x = self.classifier(x)
return x
def _initialize_weights(self):
for m in self.modules():
# if isinstance(m, nn.Conv2d) or isinstance(m, nn.Linear):
# nn.init.xavier_normal(m.weight)
# if m.bias is not None:
# nn.init.constant(m.bias, 0)
# elif isinstance(m, nn.BatchNorm2d):
# nn.init.constant(m.weight, 1)
# nn.init.constant(m.bias, 0)
if isinstance(m, nn.Conv2d):
n = m.kernel_size[0] * m.kernel_size[1] * m.out_channels
m.weight.data.normal_(0, math.sqrt(2. / n))
if m.bias is not None:
m.bias.data.zero_()
elif isinstance(m, nn.BatchNorm2d):
m.weight.data.fill_(1)
m.bias.data.zero_()
elif isinstance(m, nn.Linear):
n = m.weight.size(1)
m.weight.data.normal_(0, 0.01)
m.bias.data.zero_()