-
Notifications
You must be signed in to change notification settings - Fork 0
/
model.py
153 lines (128 loc) · 4.89 KB
/
model.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
import numpy as np
import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
def init_weights_func(model):
classname = model.__class__.__name__
if classname.find("Conv") != -1:
torch.nn.init.normal_(model.weight.data, 0.0, 0.02)
if hasattr(model, "bias") and model.bias is not None:
torch.nn.init.constant_(model.bias.data, 0.0)
### From original CycleGAN paper: https://arxiv.org/pdf/1703.10593.pdf ###
# c7s1-k denotes a 7×7 Convolution-InstanceNormReLU layer with k filters and stride 1.
# dk denotes a 3 × 3 Convolution-InstanceNorm-ReLU layer with k filters and stride 2.
# Reflection padding was used to reduce artifacts.
# Rk denotes a residual block that contains two 3 × 3 convolutional layers with the same number of filters on both layers.
# uk denotes a 3 × 3 fractional-strided-ConvolutionInstanceNorm-ReLU layer with k filters and stride 1/2.
# The network with 6 residual blocks consists of:
# c7s1-64,d128,d256,R256,R256,R256,R256,R256,R256,u128,u64,c7s1-3
###
class Generator(nn.Module):
def __init__(self, in_channels=3, num_residual=6):
super().__init__()
# c7s1-64
conv1 = nn.Sequential(
nn.ReflectionPad2d(in_channels),
nn.Conv2d(in_channels=in_channels, out_channels=64, kernel_size=7, stride=1),
nn.InstanceNorm2d(64),
nn.ReLU(inplace=True)
)
# d128
down1 = nn.Sequential(
nn.Conv2d(in_channels=64, out_channels=128, kernel_size=3, stride=2, padding=1),
nn.InstanceNorm2d(128),
nn.ReLU(inplace=True)
)
# d256
down2 = nn.Sequential(
nn.Conv2d(in_channels=128, out_channels=256, kernel_size=3, stride=2, padding=1),
nn.InstanceNorm2d(256),
nn.ReLU(inplace=True)
)
res_layer = []
for _ in range(num_residual):
# R256
res_layer += [ResBlock()]
res_layer = nn.Sequential(*res_layer)
# u128
up1 = nn.Sequential(
nn.Upsample(scale_factor=2),
nn.Conv2d(256, 128, 3, stride=1, padding=1),
nn.InstanceNorm2d(128),
nn.ReLU(inplace=True),
)
# u64
up2 = nn.Sequential(
nn.Upsample(scale_factor=2),
nn.Conv2d(128, 64, 3, stride=1, padding=1),
nn.InstanceNorm2d(64),
nn.ReLU(inplace=True),
)
#c7s1-3
conv2 = nn.Sequential(
nn.ReflectionPad2d(in_channels),
nn.Conv2d(in_channels=64, out_channels=3, kernel_size=7, stride=1),
nn.Tanh()
)
self.generator = nn.Sequential(conv1, down1, down2, res_layer, up1, up2, conv2)
def forward(self, x):
return self.generator(x)
class ResBlock(nn.Module):
def __init__(self, in_channels=256, kernel_size=3):
super().__init__()
self.layer1 = nn.Sequential(
nn.ReflectionPad2d(1),
nn.Conv2d(
in_channels=in_channels,
out_channels=in_channels,
kernel_size=kernel_size
),
nn.InstanceNorm2d(in_channels),
)
self.layer2 = nn.Sequential(
nn.ReflectionPad2d(1),
nn.Conv2d(
in_channels=in_channels,
out_channels=in_channels,
kernel_size=kernel_size
),
nn.InstanceNorm2d(in_channels),
)
def forward(self, x):
c1 = self.layer1(x)
c1 = F.relu(c1)
c2 = self.layer2(c1)
return x + c2
# For discriminator networks, we use 70 × 70 PatchGAN.
# Let Ck denote a 4 × 4 Convolution-InstanceNorm-LeakyReLU layer with k filters and stride 2.
# After the last layer, we apply a convolution to produce a 1-dimensional output.
# We do not use InstanceNorm for the first C64 layer.
# We use leaky ReLUs with a slope of 0.2.
# The discriminator architecture is: C64-C128-C256-C512
class Discriminator(nn.Module):
def __init__(self):
super().__init__()
conv1 = nn.Sequential(
nn.Conv2d(3, 64, 4, stride=2, padding=1),
nn.LeakyReLU(0.2, inplace=True)
)
conv2 = nn.Sequential(
nn.Conv2d(64, 128, 4, stride=2, padding=1),
nn.InstanceNorm2d(128),
nn.LeakyReLU(0.2, inplace=True)
)
conv3 = nn.Sequential(
nn.Conv2d(128, 256, 4, stride=2, padding=1),
nn.InstanceNorm2d(256),
nn.LeakyReLU(0.2, inplace=True)
)
conv4 = nn.Sequential(
nn.Conv2d(256, 512, 4, stride=1, padding=1),
nn.InstanceNorm2d(512),
nn.LeakyReLU(0.2, inplace=True),
nn.Conv2d(512, 1, 4, stride=1, padding=1)
)
self.discriminator = nn.Sequential(conv1, conv2, conv3, conv4)
def forward(self, x):
return self.discriminator(x)