forked from kevinlin311tw/keras-openpose-reproduce
-
Notifications
You must be signed in to change notification settings - Fork 0
/
model.py
executable file
·210 lines (159 loc) · 6.68 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
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
from keras.models import Model
from keras.layers.merge import Concatenate
from keras.layers import Activation, Input, Lambda
from keras.layers.convolutional import Conv2D
from keras.layers.pooling import MaxPooling2D
from keras.layers.merge import Multiply
from keras.regularizers import l2
from keras.initializers import random_normal, constant
stages = 6
np_branch1 = 38
np_branch2 = 19
def relu(x): return Activation('relu')(x)
def conv(x, nf, ks, name, weight_decay):
kernel_reg = l2(weight_decay[0]) if weight_decay else None
bias_reg = l2(weight_decay[1]) if weight_decay else None
x = Conv2D(nf, (ks, ks), padding='same', name=name,
kernel_regularizer=kernel_reg,
bias_regularizer=bias_reg,
kernel_initializer=random_normal(stddev=0.01),
bias_initializer=constant(0.0))(x)
return x
def pooling(x, ks, st, name):
x = MaxPooling2D((ks, ks), strides=(st, st), name=name)(x)
return x
def vgg_block(x, weight_decay):
# Block 1
x = conv(x, 64, 3, "conv1_1", (weight_decay, 0))
x = relu(x)
x = conv(x, 64, 3, "conv1_2", (weight_decay, 0))
x = relu(x)
x = pooling(x, 2, 2, "pool1_1")
# Block 2
x = conv(x, 128, 3, "conv2_1", (weight_decay, 0))
x = relu(x)
x = conv(x, 128, 3, "conv2_2", (weight_decay, 0))
x = relu(x)
x = pooling(x, 2, 2, "pool2_1")
# Block 3
x = conv(x, 256, 3, "conv3_1", (weight_decay, 0))
x = relu(x)
x = conv(x, 256, 3, "conv3_2", (weight_decay, 0))
x = relu(x)
x = conv(x, 256, 3, "conv3_3", (weight_decay, 0))
x = relu(x)
x = conv(x, 256, 3, "conv3_4", (weight_decay, 0))
x = relu(x)
x = pooling(x, 2, 2, "pool3_1")
# Block 4
x = conv(x, 512, 3, "conv4_1", (weight_decay, 0))
x = relu(x)
x = conv(x, 512, 3, "conv4_2", (weight_decay, 0))
x = relu(x)
# Additional non vgg layers
x = conv(x, 256, 3, "conv4_3_CPM", (weight_decay, 0))
x = relu(x)
x = conv(x, 128, 3, "conv4_4_CPM", (weight_decay, 0))
x = relu(x)
return x
def stage1_block(x, num_p, branch, weight_decay):
# Block 1
x = conv(x, 128, 3, "Mconv1_stage1_L%d" % branch, (weight_decay, 0))
x = relu(x)
x = conv(x, 128, 3, "Mconv2_stage1_L%d" % branch, (weight_decay, 0))
x = relu(x)
x = conv(x, 128, 3, "Mconv3_stage1_L%d" % branch, (weight_decay, 0))
x = relu(x)
x = conv(x, 512, 1, "Mconv4_stage1_L%d" % branch, (weight_decay, 0))
x = relu(x)
x = conv(x, num_p, 1, "Mconv5_stage1_L%d" % branch, (weight_decay, 0))
return x
def stageT_block(x, num_p, stage, branch, weight_decay):
# Block 1
x = conv(x, 128, 7, "Mconv1_stage%d_L%d" % (stage, branch), (weight_decay, 0))
x = relu(x)
x = conv(x, 128, 7, "Mconv2_stage%d_L%d" % (stage, branch), (weight_decay, 0))
x = relu(x)
x = conv(x, 128, 7, "Mconv3_stage%d_L%d" % (stage, branch), (weight_decay, 0))
x = relu(x)
x = conv(x, 128, 7, "Mconv4_stage%d_L%d" % (stage, branch), (weight_decay, 0))
x = relu(x)
x = conv(x, 128, 7, "Mconv5_stage%d_L%d" % (stage, branch), (weight_decay, 0))
x = relu(x)
x = conv(x, 128, 1, "Mconv6_stage%d_L%d" % (stage, branch), (weight_decay, 0))
x = relu(x)
x = conv(x, num_p, 1, "Mconv7_stage%d_L%d" % (stage, branch), (weight_decay, 0))
return x
def apply_mask(x, mask1, mask2, num_p, stage, branch):
w_name = "weight_stage%d_L%d" % (stage, branch)
if num_p == np_branch1:
w = Multiply(name=w_name)([x, mask1]) # vec_weight
elif num_p == np_branch2:
w = Multiply(name=w_name)([x, mask2]) # vec_heat
else:
assert False, "wrong number of layers num_p=%d " % num_p
return w
def get_training_model(weight_decay, gpus=None):
img_input_shape = (None, None, 3)
vec_input_shape = (None, None, 38)
heat_input_shape = (None, None, 19)
inputs = []
outputs = []
img_input = Input(shape=img_input_shape)
vec_weight_input = Input(shape=vec_input_shape)
heat_weight_input = Input(shape=heat_input_shape)
inputs.append(img_input)
inputs.append(vec_weight_input)
inputs.append(heat_weight_input)
img_normalized = Lambda(lambda x: x / 256 - 0.5)(img_input) # [-0.5, 0.5]
# VGG
stage0_out = vgg_block(img_normalized, weight_decay)
# stage 1 - branch 1 (PAF)
stage1_branch1_out = stage1_block(stage0_out, np_branch1, 1, weight_decay)
w1 = apply_mask(stage1_branch1_out, vec_weight_input, heat_weight_input, np_branch1, 1, 1)
# stage 1 - branch 2 (confidence maps)
stage1_branch2_out = stage1_block(stage0_out, np_branch2, 2, weight_decay)
w2 = apply_mask(stage1_branch2_out, vec_weight_input, heat_weight_input, np_branch2, 1, 2)
x = Concatenate()([stage1_branch1_out, stage1_branch2_out, stage0_out])
outputs.append(w1)
outputs.append(w2)
# stage sn >= 2
for sn in range(2, stages + 1):
# stage SN - branch 1 (PAF)
stageT_branch1_out = stageT_block(x, np_branch1, sn, 1, weight_decay)
w1 = apply_mask(stageT_branch1_out, vec_weight_input, heat_weight_input, np_branch1, sn, 1)
# stage SN - branch 2 (confidence maps)
stageT_branch2_out = stageT_block(x, np_branch2, sn, 2, weight_decay)
w2 = apply_mask(stageT_branch2_out, vec_weight_input, heat_weight_input, np_branch2, sn, 2)
outputs.append(w1)
outputs.append(w2)
if (sn < stages):
x = Concatenate()([stageT_branch1_out, stageT_branch2_out, stage0_out])
if gpus is None:
model = Model(inputs=inputs, outputs=outputs)
else:
import tensorflow as tf
with tf.device('/cpu:0'): #this model will not be actually used, it's template
model = Model(inputs=inputs, outputs=outputs)
return model
def get_testing_model():
img_input_shape = (None, None, 3)
img_input = Input(shape=img_input_shape)
img_normalized = Lambda(lambda x: x / 256 - 0.5)(img_input) # [-0.5, 0.5]
# VGG
stage0_out = vgg_block(img_normalized, None)
# stage 1 - branch 1 (PAF)
stage1_branch1_out = stage1_block(stage0_out, np_branch1, 1, None)
# stage 1 - branch 2 (confidence maps)
stage1_branch2_out = stage1_block(stage0_out, np_branch2, 2, None)
x = Concatenate()([stage1_branch1_out, stage1_branch2_out, stage0_out])
# stage t >= 2
stageT_branch1_out = None
stageT_branch2_out = None
for sn in range(2, stages + 1):
stageT_branch1_out = stageT_block(x, np_branch1, sn, 1, None)
stageT_branch2_out = stageT_block(x, np_branch2, sn, 2, None)
if (sn < stages):
x = Concatenate()([stageT_branch1_out, stageT_branch2_out, stage0_out])
model = Model(inputs=[img_input], outputs=[stageT_branch1_out, stageT_branch2_out])
return model