-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathmodels.py
84 lines (70 loc) · 3.86 KB
/
models.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
from keras.layers import Dense
from keras.layers import Input, Conv2D, MaxPooling2D, Flatten, Dropout, BatchNormalization
from keras.models import Model
# Default Omnisphero Model
############
def omnisphero_model(n_classes: int, input_height: int, input_width: int, input_depth: int, data_format: str) -> Model:
"""prototype model for single class decision
"""
# Input
img_input = Input(shape=(input_height, input_width, input_depth), name='input_layer')
# Convolution Blocks (FEATURE EXTRACTION)
# Conv Block 1
c1 = Conv2D(32, (3, 3), kernel_initializer='he_uniform', activation='relu', padding='same', name='block1_conv1',
data_format=data_format)(img_input)
bn1 = BatchNormalization(name='batch_norm_1')(c1)
c2 = Conv2D(32, (3, 3), kernel_initializer='he_uniform', activation='relu', padding='same', name='block1_conv2',
data_format=data_format)(bn1)
bn2 = BatchNormalization(name='batch_norm_2')(c2)
p1 = MaxPooling2D((2, 2), name='block1_pooling', data_format=data_format)(bn2)
block1 = p1
# Dave's Idee:
# #Conv Block 1
# c1 = Conv2D(32, (3,3), padding='same', name='block1_conv1', data_format=data_format)(img_input)
# bn1 = BatchNormalization(name='batch_norm_1')(c1)
# act1 = Activation('relu', alpha=0.0, max_value=None, threshold=0.0)(bn1)
#
# c2 = Conv2D(32, (3,3), activation='relu', padding='same', name='block1_conv2', data_format=data_format)(act1)
# bn2 = BatchNormalization(name='batch_norm_2')(c2)
# p1 = MaxPooling2D((2,2), name='block1_pooling', data_format=data_format)(bn2)
# block1 = p1
# Conv Block 2
c3 = Conv2D(64, (3, 3), kernel_initializer='he_uniform', activation='relu', padding='same', name='block2_conv1',
data_format=data_format)(block1)
bn3 = BatchNormalization(name='batch_norm_3')(c3)
c4 = Conv2D(64, (3, 3), kernel_initializer='he_uniform', activation='relu', padding='same', name='block2_conv2',
data_format=data_format)(bn3)
bn4 = BatchNormalization(name='batch_norm_4')(c4)
p2 = MaxPooling2D((2, 2), name='block2_pooling', data_format='channels_last')(bn4)
block2 = p2
# Conv Block 3
c5 = Conv2D(128, (3, 3), kernel_initializer='he_uniform', activation='relu', padding='same', name='block3_conv1',
data_format=data_format)(block2)
bn5 = BatchNormalization(name='batch_norm_5')(c5)
c6 = Conv2D(128, (3, 3), kernel_initializer='he_uniform', activation='relu', padding='same', name='block3_conv2',
data_format=data_format)(bn5)
bn6 = BatchNormalization(name='batch_norm_6')(c6)
p3 = MaxPooling2D((2, 2), name='block3_pooling', data_format='channels_last')(bn6)
block3 = p3
# Conv Block 4
c7 = Conv2D(256, (3, 3), kernel_initializer='he_uniform', activation='relu', padding='same', name='block4_conv1',
data_format=data_format)(block3)
bn7 = BatchNormalization(name='batch_norm_7')(c7)
c8 = Conv2D(256, (3, 3), kernel_initializer='he_uniform', activation='relu', padding='same', name='block4_conv2',
data_format=data_format)(bn7)
bn8 = BatchNormalization(name='batch_norm_8')(c8)
p4 = MaxPooling2D((2, 2), name='block4_pooling', data_format='channels_last')(bn8)
block4 = p4
# Fully-Connected Block (CLASSIFICATION)
flat = Flatten(name='flatten')(block3)
fc1 = Dense(256, kernel_initializer='he_uniform', activation='relu', name='fully_connected1')(flat)
drop_fc_1 = Dropout(0.5)(fc1)
if n_classes == 1:
prediction = Dense(n_classes, activation='sigmoid', name='output_layer')(drop_fc_1)
else:
prediction = Dense(n_classes, activation='softmax')(drop_fc_1)
# Construction
model = Model(inputs=img_input, outputs=prediction)
return model
if __name__ == "__main__":
print('Use this function to load different models, based on your preferences.')