-
Notifications
You must be signed in to change notification settings - Fork 40
/
net.py
104 lines (91 loc) · 4.31 KB
/
net.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
#!/usr/bin/env python
from __future__ import print_function
import numpy
import chainer
from chainer import cuda
import chainer.functions as F
import chainer.links as L
# U-net https://arxiv.org/pdf/1611.07004v1.pdf
# convolution-batchnormalization-(dropout)-relu
class CBR(chainer.Chain):
def __init__(self, ch0, ch1, bn=True, sample='down', activation=F.relu, dropout=False):
self.bn = bn
self.activation = activation
self.dropout = dropout
layers = {}
w = chainer.initializers.Normal(0.02)
if sample=='down':
layers['c'] = L.Convolution2D(ch0, ch1, 4, 2, 1, initialW=w)
else:
layers['c'] = L.Deconvolution2D(ch0, ch1, 4, 2, 1, initialW=w)
if bn:
layers['batchnorm'] = L.BatchNormalization(ch1)
super(CBR, self).__init__(**layers)
def __call__(self, x):
h = self.c(x)
if self.bn:
h = self.batchnorm(h)
if self.dropout:
h = F.dropout(h)
if not self.activation is None:
h = self.activation(h)
return h
class Encoder(chainer.Chain):
def __init__(self, in_ch):
layers = {}
w = chainer.initializers.Normal(0.02)
layers['c0'] = L.Convolution2D(in_ch, 64, 3, 1, 1, initialW=w)
layers['c1'] = CBR(64, 128, bn=True, sample='down', activation=F.leaky_relu, dropout=False)
layers['c2'] = CBR(128, 256, bn=True, sample='down', activation=F.leaky_relu, dropout=False)
layers['c3'] = CBR(256, 512, bn=True, sample='down', activation=F.leaky_relu, dropout=False)
layers['c4'] = CBR(512, 512, bn=True, sample='down', activation=F.leaky_relu, dropout=False)
layers['c5'] = CBR(512, 512, bn=True, sample='down', activation=F.leaky_relu, dropout=False)
layers['c6'] = CBR(512, 512, bn=True, sample='down', activation=F.leaky_relu, dropout=False)
layers['c7'] = CBR(512, 512, bn=True, sample='down', activation=F.leaky_relu, dropout=False)
super(Encoder, self).__init__(**layers)
def __call__(self, x):
hs = [F.leaky_relu(self.c0(x))]
for i in range(1,8):
hs.append(self['c%d'%i](hs[i-1]))
return hs
class Decoder(chainer.Chain):
def __init__(self, out_ch):
layers = {}
w = chainer.initializers.Normal(0.02)
layers['c0'] = CBR(512, 512, bn=True, sample='up', activation=F.relu, dropout=True)
layers['c1'] = CBR(1024, 512, bn=True, sample='up', activation=F.relu, dropout=True)
layers['c2'] = CBR(1024, 512, bn=True, sample='up', activation=F.relu, dropout=True)
layers['c3'] = CBR(1024, 512, bn=True, sample='up', activation=F.relu, dropout=False)
layers['c4'] = CBR(1024, 256, bn=True, sample='up', activation=F.relu, dropout=False)
layers['c5'] = CBR(512, 128, bn=True, sample='up', activation=F.relu, dropout=False)
layers['c6'] = CBR(256, 64, bn=True, sample='up', activation=F.relu, dropout=False)
layers['c7'] = L.Convolution2D(128, out_ch, 3, 1, 1, initialW=w)
super(Decoder, self).__init__(**layers)
def __call__(self, hs):
h = self.c0(hs[-1])
for i in range(1,8):
h = F.concat([h, hs[-i-1]])
if i<7:
h = self['c%d'%i](h)
else:
h = self.c7(h)
return h
class Discriminator(chainer.Chain):
def __init__(self, in_ch, out_ch):
layers = {}
w = chainer.initializers.Normal(0.02)
layers['c0_0'] = CBR(in_ch, 32, bn=False, sample='down', activation=F.leaky_relu, dropout=False)
layers['c0_1'] = CBR(out_ch, 32, bn=False, sample='down', activation=F.leaky_relu, dropout=False)
layers['c1'] = CBR(64, 128, bn=True, sample='down', activation=F.leaky_relu, dropout=False)
layers['c2'] = CBR(128, 256, bn=True, sample='down', activation=F.leaky_relu, dropout=False)
layers['c3'] = CBR(256, 512, bn=True, sample='down', activation=F.leaky_relu, dropout=False)
layers['c4'] = L.Convolution2D(512, 1, 3, 1, 1, initialW=w)
super(Discriminator, self).__init__(**layers)
def __call__(self, x_0, x_1):
h = F.concat([self.c0_0(x_0), self.c0_1(x_1)])
h = self.c1(h)
h = self.c2(h)
h = self.c3(h)
h = self.c4(h)
#h = F.average_pooling_2d(h, h.data.shape[2], 1, 0)
return h