/
model.py
82 lines (69 loc) · 3.28 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
from tensorflow.keras.layers import Conv2D,BatchNormalization,LeakyReLU,Dense,Flatten,Dropout,Reshape
from tensorflow import keras
import tensorflow as tf
class Res_block(keras.Model):
def __init__(self,output_plain,stride):
super(Res_block, self).__init__()
self.BN1 = BatchNormalization()
self.conv1 = Conv2D(output_plain, kernel_size=3, strides=1, padding='same',kernel_initializer='glorot_normal')
self.ac1 = LeakyReLU(0.4)
self.BN2 = BatchNormalization()
self.conv2 = Conv2D(output_plain, kernel_size=3, strides=stride, padding='same',kernel_initializer='glorot_normal')
if stride != 1:
self.downsample = tf.keras.Sequential()
self.downsample.add(Conv2D(output_plain, kernel_size=1, strides=stride, padding='same',kernel_initializer='glorot_normal'))
self.downsample.add(BatchNormalization())
def call(self, inputs, training=False, **kwargs):
res = self.downsample(inputs)
out = self.BN1(inputs)
out = self.conv1(out)
out = self.ac1(out)
out = self.BN2(out)
out = self.conv2(out)
res += out
return res
class PDSN_model(keras.Model):
def __init__(self):
super(PDSN_model, self).__init__()
### Siamese Network
self.layer0 = self.make_basic_block_layer(64, 2)
self.layer1 = self.make_basic_block_layer(128, 2)
self.layer2 = self.make_basic_block_layer(256, 2)
self.layer3 = self.make_basic_block_layer(512, 2)
### mask generator
self.conv = Conv2D(512, kernel_size=3, strides=1, padding='same', kernel_initializer='glorot_normal', activation=LeakyReLU(0.4))
self.BN1 = BatchNormalization()
### Cls for Reg mnist
self.flatten = Flatten()
self.NN1 = Dense(128, kernel_initializer='glorot_normal', activation='sigmoid')
self.dropout = Dropout(0.3)
self.NN2 = Dense(10, kernel_initializer='glorot_normal', activation='softmax')
def make_basic_block_layer(self, filter_num, stride):
return Res_block(filter_num, stride)
def Siamese_Network(self,x):
out = self.layer0(x)
out = self.layer1(out)
out = self.layer2(out)
out = self.layer3(out)
return out
def mask_generator(self,x):
out = self.conv(x)
out = self.BN1(out)
out = tf.math.sigmoid(out)
return out
def Cls(self, x):
out = self.flatten(x)
out = self.NN1(out)
out = self.dropout(out)
out = self.NN2(out)
return out
def call(self, inputs, training=False, **kwargs):
Nonocc_image, Occ_image = inputs[:,:,:,0], inputs[:,:,:,1]
Nonocc_image, Occ_image = tf.expand_dims(Nonocc_image,axis=-1), tf.expand_dims(Occ_image,axis=-1)
Nonocc_feature, Occ_feature = self.Siamese_Network(Nonocc_image), self.Siamese_Network(Occ_image)
Z_different = tf.abs(Nonocc_feature - Occ_feature)
Z_different = self.mask_generator(Z_different)
Nonocc_feature_x_Z_different = tf.multiply(Z_different,Nonocc_feature)
Occ_feature_x_Z_different = tf.multiply(Z_different, Occ_feature)
pred_label_for_Occ = self.Cls(Occ_feature_x_Z_different)
return Z_different, Nonocc_feature_x_Z_different,Occ_feature_x_Z_different,pred_label_for_Occ