-
Notifications
You must be signed in to change notification settings - Fork 30
/
model_recog_def.py
239 lines (211 loc) · 7.3 KB
/
model_recog_def.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
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
# -*- coding: utf-8 -*-
import tensorflow as tf
import zoo_layers as layers
#
# model
#
def conv_feat_layers(inputs, width, training):
#
# convolutional features maps for recognition
#
#
# recog-inputs should have shape [ batch, 36, width, channel]
#
# height_norm = 36
#
#
# [3,1; 1,1],
# [9,2; 3,2], [9,2; 3,2], [9,2; 3,2]
# [18,4; 6,4], [18,4; 6,4], [18,4; 6,4]
# [36,8; 12,8], [36,8; 12,8], [36,8; 12,8],
#
#
layer_params = [ [ 64, (3,3), (1,1), 'same', True, True, 'conv1'],
[ 64, (3,3), (1,1), 'same', True, True, 'conv2'],
[ 64, (2,2), (2,2), 'valid', True, True, 'pool1'], # for pool
[ 128, (3,3), (1,1), 'same', True, True, 'conv3'],
[ 128, (3,3), (1,1), 'same', True, True, 'conv4'],
[ 128, (2,2), (2,2), 'valid', True, True, 'pool2'], # for pool
[ 256, (3,3), (1,1), 'same', True, True, 'conv5'],
[ 256, (3,3), (1,1), 'same', True, True, 'conv6'],
[ 256, (3,2), (3,2), 'valid', True, True, 'pool3'], # for pool
[ 512, (3,1), (1,1), 'valid', True, True, 'conv_feat'] ] # for feat
#
with tf.variable_scope("conv_comm"):
#
inputs = layers.conv_layer(inputs, layer_params[0], training)
inputs = layers.conv_layer(inputs, layer_params[1], training)
inputs = layers.padd_layer(inputs, [[0,0],[0,0],[0,1],[0,0]], name='padd1')
inputs = layers.conv_layer(inputs, layer_params[2], training)
#inputs = layers.maxpool_layer(inputs, (2,2), (2,2), 'valid', 'pool1')
#
params = [[ 64, 3, (1,1), 'same', True, True, 'conv1'],
[ 64, 3, (1,1), 'same', True, False, 'conv2']]
inputs = layers.block_resnet_others(inputs, params, True, training, 'res1')
#
inputs = layers.conv_layer(inputs, layer_params[3], training)
inputs = layers.conv_layer(inputs, layer_params[4], training)
inputs = layers.padd_layer(inputs, [[0,0],[0,0],[0,1],[0,0]], name='padd2')
inputs = layers.conv_layer(inputs, layer_params[5], training)
#inputs = layers.maxpool_layer(inputs, (2,2), (2,2), 'valid', 'pool2')
#
params = [[ 128, 3, (1,1), 'same', True, True, 'conv1'],
[ 128, 3, (1,1), 'same', True, False, 'conv2']]
inputs = layers.block_resnet_others(inputs, params, True, training, 'res2')
#
inputs = layers.conv_layer(inputs, layer_params[6], training)
inputs = layers.conv_layer(inputs, layer_params[7], training)
inputs = layers.padd_layer(inputs, [[0,0],[0,0],[0,1],[0,0]], name='padd3')
inputs = layers.conv_layer(inputs, layer_params[8], training)
#inputs = layers.maxpool_layer(inputs, (3,2), (3,2), 'valid', 'pool3')
#
params = [[ 256, 3, (1,1), 'same', True, True, 'conv1'],
[ 256, 3, (1,1), 'same', True, False, 'conv2']]
inputs = layers.block_resnet_others(inputs, params, True, training, 'res3')
#
conv_feat = layers.conv_layer(inputs, layer_params[9], training)
#
#
# Calculate resulting sequence length from original image widths
#
two = tf.constant(2, dtype=tf.float32, name='two')
#
w = tf.cast(width, tf.float32)
#
w = tf.div(w, two)
w = tf.ceil(w)
#
w = tf.div(w, two)
w = tf.ceil(w)
#
w = tf.div(w, two)
w = tf.ceil(w)
#
w = tf.cast(w, tf.int32)
#
# Vectorize
sequence_length = tf.reshape(w, [-1], name='seq_len')
#
#
return conv_feat, sequence_length
#
#
def rnn_recog_layers(features, sequence_length, num_classes):
#
# batch-picture features
features = tf.squeeze(features, axis = 1) # squeeze
#
# [batchSize paddedSeqLen numFeatures]
#
#
rnn_size = 256 # 256, 512
#fc_size = 512 # 256, 384, 512
#
weight_initializer = tf.contrib.layers.variance_scaling_initializer()
bias_initializer = tf.constant_initializer(value=0.0)
#
#
# Transpose to time-major order for efficiency
# --> [paddedSeqLen batchSize numFeatures]
#
rnn_sequence = tf.transpose(features, perm = [1, 0, 2], name = 'time_major')
#
rnn1 = layers.gru_layer(rnn_sequence, sequence_length, rnn_size, 'bdrnn1')
rnn2 = layers.gru_layer(rnn1, sequence_length, rnn_size, 'bdrnn2')
#
# out
#
rnn_logits = tf.layers.dense(rnn2, num_classes,
activation = None, #tf.nn.sigmoid,
kernel_initializer = weight_initializer,
bias_initializer = bias_initializer,
name = 'rnn_logits')
#
# dense operates on last dim
#
#
return rnn_logits
#
def ctc_loss_layer(sequence_labels, rnn_logits, sequence_length):
#
loss = tf.nn.ctc_loss(inputs = rnn_logits,
labels = sequence_labels,
sequence_length = sequence_length,
ignore_longer_outputs_than_inputs = True,
time_major = True )
#
total_loss = tf.reduce_mean(loss, name = 'loss')
#
return total_loss
#
def decode_rnn_results_ctc_beam(results, seq_len):
#
# tf.nn.ctc_beam_search_decoder
#
decoded, log_prob = tf.nn.ctc_beam_search_decoder(results, seq_len, merge_repeated=False)
#
return decoded
#
'''
sequence_length: 1-D int32 vector, size [batch_size].The sequence lengths.
'''
'''
Example: The sparse tensor
SparseTensor(values=[1, 2], indices=[[0, 0], [1, 2]], dense_shape=[3, 4])
represents the dense tensor
[[1, 0, 0, 0]
[0, 0, 2, 0]
[0, 0, 0, 0]]
'''
#
# labels: An int32 SparseTensor.
#
# labels.indices[i, :] == [b, t] means
# labels.values[i] stores the id for (batch b, time t).
# labels.values[i] must take on values in [0, num_labels).
#
# sparse_targets = sparse_tuple_from(targets)
#
def convert2SparseTensorValue(list_labels):
#
# list_labels: batch_major
#
#
num_samples = len(list_labels)
num_maxlen = max(map(lambda x: len(x), list_labels))
#
indices = []
values = []
shape = [num_samples, num_maxlen]
#
for idx in range(num_samples):
#
item = list_labels[idx]
#
values.extend(item)
indices.extend([[idx, posi] for posi in range(len(item))])
#
#
return tf.SparseTensorValue(indices = indices, values = values, dense_shape = shape)
#
#
def convert2ListLabels(sparse_tensor_value):
#
# list_labels: batch_major
#
shape = sparse_tensor_value.dense_shape
indices = sparse_tensor_value.indices
values = sparse_tensor_value.values
list_labels = []
#
item = [0]*shape[1]
for i in range(shape[0]): list_labels.append(item)
#
for idx, value in enumerate(values):
#
posi = indices[idx]
#
list_labels[posi[0]][posi[1]] = value
#
return list_labels
#