-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy path0801_mel_spectrogram_crnn.py
193 lines (146 loc) · 7.86 KB
/
0801_mel_spectrogram_crnn.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
# -*- coding:utf-8 -*-
"""
@author: Songgx
@file: 0801_mel_spectrogram_crnn.py
@time: 2017/2/18 15:42
"""
import tensorflow as tf
import sklearn.metrics as sm
import numpy as np
batch_size = 20
learning_rate = 0.003
n_epoch = 400
n_classes = 10 # total classes (0-9 digits)
x_height = 96
x_width = 1366
def init_weights(shape):
return tf.Variable(tf.random_normal(shape, stddev=0.01))
def init_biases(shape):
return tf.Variable(tf.zeros(shape))
def batch_norm(x, n_out, phase_train, scope='bn'):
with tf.variable_scope(scope):
beta = tf.Variable(tf.constant(0.0, shape=[n_out]), name='beta', trainable=True)
gamma = tf.Variable(tf.constant(1.0, shape=[n_out]), name='gamma', trainable=True)
batch_mean, batch_var = tf.nn.moments(x, [0, 1, 2], name='moments')
ema = tf.train.ExponentialMovingAverage(decay=0.5)
def mean_var_with_update():
ema_apply_op = ema.apply([batch_mean, batch_var])
with tf.control_dependencies([ema_apply_op]):
return tf.identity(batch_mean), tf.identity(batch_var)
mean, var = tf.cond(phase_train,
mean_var_with_update,
lambda: (ema.average(batch_mean), ema.average(batch_var)))
normed = tf.nn.batch_normalization(x, mean, var, beta, gamma, 1e-3)
return normed
def read_and_decode(filename):
filename_queue = tf.train.string_input_producer([filename])
reader = tf.TFRecordReader()
_, serialized_example = reader.read(filename_queue)
features = tf.parse_single_example(serialized_example,
features={
'label': tf.FixedLenFeature([n_classes], tf.float32),
'features_mel': tf.FixedLenFeature([], tf.string),
})
x = tf.decode_raw(features['features_mel'], tf.float32)
x = tf.reshape(x, [x_height,x_width, 1]) # 与placeholder 保持一致
y = tf.cast(features['label'], tf.float32)
return x, y
def crnn(melspectrogram, weights, phase_train):
x = tf.cast(tf.pad(melspectrogram, [[0, 0], [0, 0], [37, 37], [0, 0]], 'CONSTANT'), tf.float32)
x = batch_norm(tf.reshape(x, [-1, 1, 96, 1440]), 1440, phase_train)
x = tf.reshape(x, [-1, 96, 1440, 1])
conv2_1 = tf.add(tf.nn.conv2d(x, weights['wconv1'], strides=[1, 1, 1, 1], padding='SAME'), weights['bconv1'])
conv2_1 = tf.nn.relu(batch_norm(conv2_1, 64, phase_train))
mpool_1 = tf.nn.max_pool(conv2_1, ksize=[1, 2, 2, 1], strides=[1, 2, 2, 1], padding='VALID')
dropout_1 = tf.nn.dropout(mpool_1, 0.5)
conv2_2 = tf.add(tf.nn.conv2d(dropout_1, weights['wconv2'], strides=[1, 1, 1, 1], padding='SAME'),
weights['bconv2'])
conv2_2 = tf.nn.relu(batch_norm(conv2_2, 128, phase_train))
mpool_2 = tf.nn.max_pool(conv2_2, ksize=[1, 3, 3, 1], strides=[1, 3, 3, 1], padding='VALID')
dropout_2 = tf.nn.dropout(mpool_2, 0.5)
conv2_3 = tf.add(tf.nn.conv2d(dropout_2, weights['wconv3'], strides=[1, 1, 1, 1], padding='SAME'),
weights['bconv3'])
conv2_3 = tf.nn.relu(batch_norm(conv2_3, 128, phase_train))
mpool_3 = tf.nn.max_pool(conv2_3, ksize=[1, 4, 4, 1], strides=[1, 4, 4, 1], padding='VALID')
dropout_3 = tf.nn.dropout(mpool_3, 0.5)
conv2_4 = tf.add(tf.nn.conv2d(dropout_3, weights['wconv4'], strides=[1, 1, 1, 1], padding='SAME'),
weights['bconv4'])
conv2_4 = tf.nn.relu(batch_norm(conv2_4, 128, phase_train))
mpool_4 = tf.nn.max_pool(conv2_4, ksize=[1, 4, 4, 1], strides=[1, 4, 4, 1], padding='VALID')
dropout_4 = tf.nn.dropout(mpool_4, 0.5)
gru1_in = tf.reshape(dropout_4, [-1, 15, 128])
gru1 = tf.nn.rnn_cell.MultiRNNCell([tf.nn.rnn_cell.GRUCell(32)] * 15)
gru1_out, state = tf.nn.dynamic_rnn(gru1, gru1_in, dtype=tf.float32, scope='gru1')
gru2 = tf.nn.rnn_cell.MultiRNNCell([tf.nn.rnn_cell.GRUCell(32)] * 15)
gru2_out, state = tf.nn.dynamic_rnn(gru2, gru1_out, dtype=tf.float32, scope='gru2')
gru2_out = tf.transpose(gru2_out, [1, 0, 2])
gru2_out = tf.gather(gru2_out, int(gru2_out.get_shape()[0]) - 1)
dropout_5 = tf.nn.dropout(gru2_out, 0.3)
flat = tf.reshape(dropout_5, [-1, weights['woutput'].get_shape().as_list()[0]])
p_y_X = tf.nn.sigmoid(tf.add(tf.matmul(flat, weights['woutput']), weights['boutput']))
return p_y_X
def sort_result(tags, preds):
result = zip(tags, preds)
sorted_result = sorted(result, key=lambda x: x[1], reverse=True)
return [(name, '%5.3f' % score) for name, score in sorted_result]
if __name__ == '__main__':
features, label = read_and_decode("data/merge/mel_data_training.tfrecords")
features_test, label_test = read_and_decode("data/merge/mel_data_test.tfrecords")
# 使用shuffle_batch可以随机打乱输入
audio_batch, label_batch = tf.train.shuffle_batch([features, label],
batch_size=batch_size, capacity=2000,
min_after_dequeue=1000)
audio_batch_test, label_batch_test = tf.train.shuffle_batch([features_test, label_test],
batch_size=batch_size, capacity=2000,
min_after_dequeue=1000)
weights = {
'wconv1': init_weights([3, 3, 1, 64]),
'wconv2': init_weights([3, 3, 64, 128]),
'wconv3': init_weights([3, 3, 128, 128]),
'wconv4': init_weights([3, 3, 128, 128]),
'bconv1': init_biases([64]),
'bconv2': init_biases([128]),
'bconv3': init_biases([128]),
'bconv4': init_biases([128]),
'woutput': init_weights([32, 10]),
'boutput': init_biases([10])}
X = tf.placeholder("float", [None, 96, 1366, 1])
y = tf.placeholder("float", [None, 10])
lrate = tf.placeholder("float")
phase_train = np.array(True)
y_ = crnn(X, weights, phase_train)
cost = tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(y_, y))
train_op = tf.train.AdamOptimizer(learning_rate=learning_rate).minimize(cost)
predict_op = y_
init = tf.global_variables_initializer()
with tf.Session() as sess:
sess.run(init)
# Start input enqueue threads.
# 缺少这2行， audio_batch_vals, label_batch_vals = sess.run([audio_batch, label_batch]) 程序会假死
coord = tf.train.Coordinator()
threads = tf.train.start_queue_runners(sess=sess, coord=coord)
for i in range(n_epoch):
audio_batch_vals, label_batch_vals = sess.run([audio_batch, label_batch])
train_input_dict = {X: audio_batch_vals,
y: label_batch_vals,
lrate: learning_rate
#phase_train: True,
}
epoch_logits, epoch_cost, _ = sess.run([y_, cost, train_op], feed_dict=train_input_dict)
print('Training Epoch : ', i+1, 'cost : ', epoch_cost)
print(epoch_logits)
#print sort_result(tags, predictions)[:5])
# test model
test_epochs = int(200/batch_size)
final_roc_auc_score = 0.
for j in range(test_epochs):
audio_test_vals, label_test_vals = sess.run([audio_batch_test, label_batch_test])
test_input_dict = {X: audio_test_vals,
y: label_test_vals
# phase_train: True,
}
predictions = sess.run(predict_op, feed_dict=test_input_dict)
cur_epoch_roc_auc_score = sm.roc_auc_score(label_test_vals, predictions, average='samples')
final_roc_auc_score += cur_epoch_roc_auc_score
print('Test Epoch : ', j+1, 'AUC : ', cur_epoch_roc_auc_score)
print('Test Auc is : ', final_roc_auc_score/test_epochs)