Repository navigation
Expand file tree
/
Copy pathnetwork.py
More file actions
296 lines (274 loc) · 12.4 KB
/
Copy pathnetwork.py
File metadata and controls
296 lines (274 loc) · 12.4 KB
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
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
import os
import numpy as np
import tensorflow as tf
from utils.data_reader import H5DataLoader, H53DDataLoader
from utils.img_utils import imsave
from utils import ops
"""
This module builds a standard U-NET for semantic segmentation.
If want VAE using pixelDCL, please visit this code:
https://github.com/HongyangGao/UVAE
"""
class PixelDCN(object):
def __init__(self, sess, conf):
self.sess = sess
self.conf = conf
self.def_params()
if not os.path.exists(conf.modeldir):
os.makedirs(conf.modeldir)
if not os.path.exists(conf.logdir):
os.makedirs(conf.logdir)
if not os.path.exists(conf.sampledir):
os.makedirs(conf.sampledir)
self.configure_networks()
self.train_summary = self.config_summary('train')
self.valid_summary = self.config_summary('valid')
def def_params(self):
self.data_format = 'NHWC'
if self.conf.data_type == '3D':
self.conv_size = (3, 3, 3)
self.pool_size = (2, 2, 2)
self.axis, self.channel_axis = (1, 2, 3), 4
self.input_shape = [
self.conf.batch, self.conf.depth, self.conf.height,
self.conf.width, self.conf.channel]
self.output_shape = [
self.conf.batch, self.conf.depth, self.conf.height,
self.conf.width]
else:
self.conv_size = (3, 3)
self.pool_size = (2, 2)
self.axis, self.channel_axis = (1, 2), 3
self.input_shape = [
self.conf.batch, self.conf.height, self.conf.width,
self.conf.channel]
self.output_shape = [
self.conf.batch, self.conf.height, self.conf.width]
def configure_networks(self):
self.build_network()
optimizer = tf.train.AdamOptimizer(self.conf.learning_rate)
self.train_op = optimizer.minimize(self.loss_op, name='train_op')
tf.set_random_seed(self.conf.random_seed)
self.sess.run(tf.global_variables_initializer())
trainable_vars = tf.trainable_variables()
self.saver = tf.train.Saver(var_list=trainable_vars, max_to_keep=0)
self.writer = tf.summary.FileWriter(self.conf.logdir, self.sess.graph)
def build_network(self):
self.inputs = tf.placeholder(
tf.float32, self.input_shape, name='inputs')
self.labels = tf.placeholder(
tf.int64, self.output_shape, name='labels')
self.predictions = self.inference(self.inputs)
self.cal_loss()
def cal_loss(self):
one_hot_labels = tf.one_hot(
self.labels, depth=self.conf.class_num,
axis=self.channel_axis, name='labels/one_hot')
losses = tf.losses.softmax_cross_entropy(
one_hot_labels, self.predictions, scope='loss/losses')
self.loss_op = tf.reduce_mean(losses, name='loss/loss_op')
self.decoded_preds = tf.argmax(
self.predictions, self.channel_axis, name='accuracy/decode_pred')
correct_prediction = tf.equal(
self.labels, self.decoded_preds,
name='accuracy/correct_pred')
self.accuracy_op = tf.reduce_mean(
tf.cast(correct_prediction, tf.float32, name='accuracy/cast'),
name='accuracy/accuracy_op')
# weights = tf.cast(
# tf.greater(self.decoded_preds, 0, name='m_iou/greater'),
# tf.int32, name='m_iou/weights')
weights = tf.cast(
tf.less(self.labels, self.conf.channel, name='m_iou/greater'),
tf.int64, name='m_iou/weights')
labels = tf.multiply(self.labels, weights, name='m_iou/mul')
self.m_iou, self.miou_op = tf.metrics.mean_iou(
self.labels, self.decoded_preds, self.conf.class_num,
weights, name='m_iou/m_ious')
def config_summary(self, name):
summarys = []
summarys.append(tf.summary.scalar(name+'/loss', self.loss_op))
summarys.append(tf.summary.scalar(name+'/accuracy', self.accuracy_op))
if name == 'valid' and self.conf.data_type == '2D':
summarys.append(
tf.summary.image(name+'/input', self.inputs, max_outputs=100))
summarys.append(
tf.summary.image(
name+'/annotation',
tf.cast(tf.expand_dims(self.labels, -1),
tf.float32), max_outputs=100))
summarys.append(
tf.summary.image(
name+'/prediction',
tf.cast(tf.expand_dims(self.decoded_preds, -1),
tf.float32), max_outputs=100))
summary = tf.summary.merge(summarys)
return summary
def inference(self, inputs):
outputs = inputs
down_outputs = []
for layer_index in range(self.conf.network_depth-1):
is_first = True if not layer_index else False
name = 'down%s' % layer_index
outputs = self.build_down_block(
outputs, name, down_outputs, is_first)
outputs = self.build_bottom_block(outputs, 'bottom')
for layer_index in range(self.conf.network_depth-2, -1, -1):
is_final = True if layer_index == 0 else False
name = 'up%s' % layer_index
down_inputs = down_outputs[layer_index]
outputs = self.build_up_block(
outputs, down_inputs, name, is_final)
return outputs
def build_down_block(self, inputs, name, down_outputs, first=False):
out_num = self.conf.start_channel_num if first else 2 * \
inputs.shape[self.channel_axis].value
conv1 = ops.conv(inputs, out_num, self.conv_size,
name+'/conv1', self.conf.data_type)
conv2 = ops.conv(conv1, out_num, self.conv_size,
name+'/conv2', self.conf.data_type)
down_outputs.append(conv2)
pool = ops.pool(conv2, self.pool_size, name +
'/pool', self.conf.data_type)
return pool
def build_bottom_block(self, inputs, name):
out_num = inputs.shape[self.channel_axis].value
conv1 = ops.conv(
inputs, 2*out_num, self.conv_size, name+'/conv1',
self.conf.data_type)
conv2 = ops.conv(
conv1, out_num, self.conv_size, name+'/conv2', self.conf.data_type)
return conv2
def build_up_block(self, inputs, down_inputs, name, final=False):
out_num = inputs.shape[self.channel_axis].value
conv1 = self.deconv_func()(
inputs, out_num, self.conv_size, name+'/conv1',
self.conf.data_type, action=self.conf.action)
conv1 = tf.concat(
[conv1, down_inputs], self.channel_axis, name=name+'/concat')
conv2 = self.conv_func()(
conv1, out_num, self.conv_size, name+'/conv2', self.conf.data_type)
out_num = self.conf.class_num if final else out_num/2
conv3 = ops.conv(
conv2, out_num, self.conv_size, name+'/conv3', self.conf.data_type,
not final)
return conv3
def deconv_func(self):
return getattr(ops, self.conf.deconv_name)
def conv_func(self):
return getattr(ops, self.conf.conv_name)
def save_summary(self, summary, step):
print('---->summarizing', step)
self.writer.add_summary(summary, step)
def train(self):
if self.conf.reload_step > 0:
self.reload(self.conf.reload_step)
if self.conf.data_type == '2D':
train_reader = H5DataLoader(
self.conf.data_dir+self.conf.train_data)
valid_reader = H5DataLoader(
self.conf.data_dir+self.conf.valid_data)
else:
train_reader = H53DDataLoader(
self.conf.data_dir+self.conf.train_data, self.input_shape)
valid_reader = H53DDataLoader(
self.conf.data_dir+self.conf.valid_data, self.input_shape)
for epoch_num in range(self.conf.max_step+1):
if epoch_num and epoch_num % self.conf.test_interval == 0:
inputs, labels = valid_reader.next_batch(self.conf.batch)
feed_dict = {self.inputs: inputs,
self.labels: labels}
loss, summary = self.sess.run(
[self.loss_op, self.valid_summary], feed_dict=feed_dict)
self.save_summary(summary, epoch_num+self.conf.reload_step)
print('----testing loss', loss)
if epoch_num and epoch_num % self.conf.summary_interval == 0:
inputs, labels = train_reader.next_batch(self.conf.batch)
feed_dict = {self.inputs: inputs,
self.labels: labels}
loss, _, summary = self.sess.run(
[self.loss_op, self.train_op, self.train_summary],
feed_dict=feed_dict)
self.save_summary(summary, epoch_num+self.conf.reload_step)
else:
inputs, labels = train_reader.next_batch(self.conf.batch)
feed_dict = {self.inputs: inputs,
self.labels: labels}
loss, _ = self.sess.run(
[self.loss_op, self.train_op], feed_dict=feed_dict)
print('----training loss', loss)
if epoch_num and epoch_num % self.conf.save_interval == 0:
self.save(epoch_num+self.conf.reload_step)
def test(self):
print('---->testing ', self.conf.test_step)
if self.conf.test_step > 0:
self.reload(self.conf.test_step)
else:
print("please set a reasonable test_step")
return
if self.conf.data_type == '2D':
test_reader = H5DataLoader(
self.conf.data_dir+self.conf.test_data, False)
else:
test_reader = H53DDataLoader(
self.conf.data_dir+self.conf.test_data, self.input_shape)
self.sess.run(tf.local_variables_initializer())
count = 0
losses = []
accuracies = []
m_ious = []
while True:
inputs, labels = test_reader.next_batch(self.conf.batch)
if inputs.shape[0] < self.conf.batch:
break
feed_dict = {self.inputs: inputs, self.labels: labels}
loss, accuracy, m_iou, _ = self.sess.run(
[self.loss_op, self.accuracy_op, self.m_iou, self.miou_op],
feed_dict=feed_dict)
print('values----->', loss, accuracy, m_iou)
count += 1
losses.append(loss)
accuracies.append(accuracy)
m_ious.append(m_iou)
print('Loss: ', np.mean(losses))
print('Accuracy: ', np.mean(accuracies))
print('M_iou: ', m_ious[-1])
def predict(self):
print('---->predicting ', self.conf.test_step)
if self.conf.test_step > 0:
self.reload(self.conf.test_step)
else:
print("please set a reasonable test_step")
return
if self.conf.data_type == '2D':
test_reader = H5DataLoader(
self.conf.data_dir+self.conf.test_data, False)
else:
test_reader = H53DDataLoader(
self.conf.data_dir+self.conf.test_data, self.input_shape)
predictions = []
while True:
inputs, labels = test_reader.next_batch(self.conf.batch)
if inputs.shape[0] < self.conf.batch:
break
feed_dict = {self.inputs: inputs, self.labels: labels}
predictions.append(self.sess.run(
self.decoded_preds, feed_dict=feed_dict))
print('----->saving predictions')
for index, prediction in enumerate(predictions):
for i in range(prediction.shape[0]):
imsave(prediction[i], self.conf.sampledir +
str(index*prediction.shape[0]+i)+'.png')
def save(self, step):
print('---->saving', step)
checkpoint_path = os.path.join(
self.conf.modeldir, self.conf.model_name)
self.saver.save(self.sess, checkpoint_path, global_step=step)
def reload(self, step):
checkpoint_path = os.path.join(
self.conf.modeldir, self.conf.model_name)
model_path = checkpoint_path+'-'+str(step)
if not os.path.exists(model_path+'.meta'):
print('------- no such checkpoint', model_path)
return
self.saver.restore(self.sess, model_path)