-
Notifications
You must be signed in to change notification settings - Fork 27
/
Net.py
165 lines (144 loc) · 6.81 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
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
import tensorflow as tf
from tensorflow.python import pywrap_tensorflow
import numpy as np
WEIGHT_INIT_STDDEV = 0.05
n = 44
class Generator(object):
def __init__(self, sco):
self.encoder = Encoder(sco)
self.decoder = Decoder(sco)
self.var_list = []
self.features = []
def transform(self, I1, I2, is_training, reuse):
img = tf.concat([I1, I2], 3)
code = self.encoder.encode(img, is_training, reuse)
generated_img = self.decoder.decode(code, is_training, reuse)
# self.var_list.extend(self.encoder.var_list)
# self.var_list.extend(self.decoder.var_list)
# self.var_list.extend(tf.trainable_variables())
return generated_img
class Encoder(object):
def __init__(self, scope_name):
self.scope = scope_name
self.var_list = []
self.weight_vars = []
with tf.variable_scope(self.scope):
with tf.variable_scope('encoder'):
self.weight_vars.append(self._create_variables(2, n, 3, scope = 'conv1_1'))
self.weight_vars.append(self._create_variables(n, n, 3, scope = 'dense_block_conv1'))
self.weight_vars.append(self._create_variables(n*2, n, 3, scope = 'dense_block_conv2'))
self.weight_vars.append(self._create_variables(n*3, n, 3, scope = 'dense_block_conv3'))
self.weight_vars.append(self._create_variables(n*4, n, 3, scope = 'dense_block_conv4'))
self.weight_vars.append(self._create_variables(n * 5, n, 3, scope = 'dense_block_conv5'))
def _create_variables(self, input_filters, output_filters, kernel_size, scope):
shape = [kernel_size, kernel_size, input_filters, output_filters]
with tf.variable_scope(scope):
kernel = tf.Variable(tf.truncated_normal(shape, stddev = WEIGHT_INIT_STDDEV),
name = 'kernel')
bias = tf.Variable(tf.zeros([output_filters]), name = 'bias')
self.var_list.append(kernel)
self.var_list.append(bias)
return (kernel, bias)
def encode(self, image, is_training, reuse):
dense_indices = [1, 2, 3, 4, 5]
out = image
for i in range(len(self.weight_vars)):
kernel, bias = self.weight_vars[i]
if i in dense_indices:
out = conv2d(out, kernel, bias, dense = True, use_lrelu = True, is_training = is_training, reuse = reuse,
Scope = self.scope + '/encoder/b' + str(i))
else:
out = conv2d(out, kernel, bias, dense = False, use_lrelu = True, is_training = is_training,
reuse = reuse, Scope = self.scope + '/encoder/b' + str(i))
return out
class Decoder(object):
def __init__(self, scope_name):
self.weight_vars = []
self.var_list = []
self.scope = scope_name
with tf.name_scope(scope_name):
with tf.variable_scope('decoder'):
self.weight_vars.append(self._create_variables(n*6, 128, 3, scope = 'conv2_1'))
self.weight_vars.append(self._create_variables(128, 64, 3, scope = 'conv2_2'))
self.weight_vars.append(self._create_variables(64, 32, 3, scope = 'conv2_3'))
self.weight_vars.append(self._create_variables(32, 1, 3, scope = 'conv2_4'))
# self.weight_vars.append(self._create_variables(32, 1, 3, scope = 'conv2_4'))
def _create_variables(self, input_filters, output_filters, kernel_size, scope):
with tf.variable_scope(scope):
shape = [kernel_size, kernel_size, input_filters, output_filters]
kernel = tf.Variable(tf.truncated_normal(shape, stddev = WEIGHT_INIT_STDDEV), name = 'kernel')
bias = tf.Variable(tf.zeros([output_filters]), name = 'bias')
self.var_list.append(kernel)
self.var_list.append(bias)
return (kernel, bias)
def decode(self, image, is_training, reuse):
final_layer_idx = len(self.weight_vars) - 1
out = image
for i in range(len(self.weight_vars)):
kernel, bias = self.weight_vars[i]
if i == final_layer_idx:
out = conv2d(out, kernel, bias, dense = False, use_lrelu = False,
Scope = self.scope + '/decoder/b' + str(i), is_training = is_training, reuse=reuse)
out = tf.nn.tanh(out) / 2 + 0.5
else:
out = conv2d(out, kernel, bias, dense = False, use_lrelu = True,
Scope = self.scope + '/decoder/b' + str(i), is_training = is_training, reuse=reuse)
return out
# class weightnet(object):
# def __init__(self, scope_name):
# self.weight_vars = []
# self.var_list = []
# self.scope = scope_name
# with tf.name_scope(scope_name):
# with tf.variable_scope('weightnet'):
# self.weight_vars.append(self._create_variables(2, 32, 3, scope = 'conv1'))
# self.weight_vars.append(self._create_variables(32, 64, 3, scope = 'conv2'))
# self.weight_vars.append(self._create_variables(64, 96, 3, scope = 'conv3'))
# self.weight_vars.append(self._create_variables(96, 2, 3, scope = 'conv4'))
#
# def _create_variables(self, input_filters, output_filters, kernel_size, scope):
# with tf.variable_scope(scope):
# shape = [kernel_size, kernel_size, input_filters, output_filters]
# kernel = tf.Variable(tf.truncated_normal(shape, stddev = WEIGHT_INIT_STDDEV), name = 'kernel')
# bias = tf.Variable(tf.zeros([output_filters]), name = 'bias')
# self.var_list.append(kernel)
# self.var_list.append(bias)
# return (kernel, bias)
#
# def generate_weight(self, image, is_training):
# final_layer_idx = len(self.weight_vars) - 1
#
# out = image
# for i in range(len(self.weight_vars)):
# kernel, bias = self.weight_vars[i]
# if i == final_layer_idx:
# out = conv2d(out, kernel, bias, use_relu = False,
# Scope = self.scope + '/generate_weight/b' + str(i), BN = False, stride=2, is_training = is_training, reuse=reuse)
# out = tf.reduce_mean(out, [1, 2], name='pool', keep_dims=True)
# b, _, _, _ = out.shape
# out = tf.reshape(out, [int(b), 2])
# out = tf.nn.softmax(out)
# else:
# out = conv2d(out, kernel, bias, use_relu = True, BN = True, stride=2,
# Scope = self.scope + '/generate_weight/b' + str(i), is_training = is_training, reuse=reuse)
# # print("%s: out shape:" % i, out.shape)
# return out
def conv2d(x, kernel, bias, use_lrelu = True, dense = False, Scope = None, stride = 1, is_training = False, reuse=False):
# padding image with reflection mode
x_padded = tf.pad(x, [[0, 0], [1, 1], [1, 1], [0, 0]], mode = 'REFLECT')
# conv and add bias
out = tf.nn.conv2d(input = x_padded, filter = kernel, strides = [1, stride, stride, 1], padding = 'VALID')
out = tf.nn.bias_add(out, bias)
# if BN:
# with tf.variable_scope(Scope):
# # print("Scope", Scope)
# # print("reuse", not is_training)
# # out = tf.contrib.layers.batch_norm(out, decay = 0.9, updates_collections = None, epsilon = 1e-5, scale = True, reuse = reuse)
#
# out = tf.layers.batch_normalization(out, training = is_training, reuse= reuse, trainable=is_training)
if use_lrelu:
# out = tf.nn.relu(out)
out = tf.maximum(out, 0.2 * out)
if dense:
out = tf.concat([out, x], 3)
return out