-
Notifications
You must be signed in to change notification settings - Fork 48
/
speech_model_12.py
223 lines (191 loc) · 11.1 KB
/
speech_model_12.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
#-*- coding:utf-8 -*-
#author:zhangwei
'''
网络模型结构为32-32-64-64-128-128-512-1422：顶层为CTC的网络结构
'''
from general_function.file_wav import *
from general_function.file_wav import *
from general_function.file_dict import *
from general_function.feature_extract import *
from general_function.edit_distance import *
import keras as kr
import numpy as np
import random
from keras.utils import plot_model
from keras.models import Model
from keras.layers import Dense , Dropout , Input , Reshape , multiply
from keras.layers import Conv2D , MaxPooling2D , Lambda , Activation , regularizers
from keras.layers.normalization import BatchNormalization
from keras import backend as K
from keras.optimizers import SGD , Adadelta , Adam
from readdata_12 import DataSpeech
class ModelSpeech():
def __init__(self , datapath):
MS_OUTPUT_SIZE = 1422
self.MS_OUTPUT_SIZE = MS_OUTPUT_SIZE
self.label_max_string_length = 64
self.AUDIO_LENGTH = 1600
self.AUDIO_FEATURE_LENGTH = 200
self.datapath = datapath
self._model , self.base_model = self.creat_model()
self.slash = '/'
if self.datapath[-1] != self.slash:
self.datapath = self.datapath + self.slash
pass
def creat_model(self):
input_data = Input(shape=[self.AUDIO_LENGTH , self.AUDIO_FEATURE_LENGTH , 1] , name='Input')
conv1 = Conv2D(filters=32 , kernel_size=[3,3] , padding='same' , activation='relu' , use_bias=True , kernel_initializer='he_normal' , kernel_regularizer=regularizers.l2(1e-4))(input_data)
conv1 = BatchNormalization(epsilon=0.0002)(conv1)
conv2 = Conv2D(filters=32 , kernel_size=[3,3] , padding='same' , activation='relu' , use_bias=True , kernel_initializer='he_normal' , kernel_regularizer=regularizers.l2(1e-4))(conv1)
conv2 = BatchNormalization(epsilon=0.0002)(conv2)
maxpool1 = MaxPooling2D(pool_size=[2,2] , strides=None , padding='valid')(conv2)
maxpool1 = Dropout(0.1)(maxpool1)
conv3 = Conv2D(filters=64 , kernel_size=[3,3] , padding='same' , activation='relu' , use_bias=True , kernel_initializer='he_normal' , kernel_regularizer=regularizers.l2(1e-4))(maxpool1)
conv3 = BatchNormalization(epsilon=0.0002)(conv3)
conv4 = Conv2D(filters=64 , kernel_size=[3,3] , padding='same' , activation='relu' , use_bias=True , kernel_initializer='he_normal' , kernel_regularizer=regularizers.l2(1e-4))(conv3)
conv4 = BatchNormalization(epsilon=0.0002)(conv4)
maxpool2 = MaxPooling2D(pool_size=[2,2] , strides=None , padding='valid')(conv4)
maxpool2 = Dropout(0.1)(maxpool2)
conv5 = Conv2D(filters=128 , kernel_size=[3,3] , padding='same' , activation='relu' , use_bias=True , kernel_initializer='he_normal' , kernel_regularizer=regularizers.l2(1e-4))(maxpool2)
conv5 = BatchNormalization(epsilon=0.0002)(conv5)
conv6 = Conv2D(filters=128 , kernel_size=[3,3] , padding='same' , activation='relu' , use_bias=True , kernel_initializer='he_normal' , kernel_regularizer=regularizers.l2(1e-4))(conv5)
conv6 = BatchNormalization(epsilon=0.0002)(conv6)
maxpool3 = MaxPooling2D(pool_size=[2,2] , strides=None , padding='valid')(conv6)
maxpool3 = Dropout(0.1)(maxpool3)
# conv7 = Conv2D(filters=128 , kernel_size=[3, 3], padding='same', activation='relu', use_bias=True , kernel_initializer='he_normal' , kernel_regularizer=regularizers.l2(1e-4))(maxpool3)
# conv7 = BatchNormalization(epsilon=0.0002)(conv7)
# conv8 = Conv2D(filters=128 , kernel_size=[3, 3], padding='same', activation='relu', use_bias=True , kernel_initializer='he_normal' , kernel_regularizer=regularizers.l2(1e-5))(conv7)
# conv8 = BatchNormalization(epsilon=0.0002)(conv8)
# maxpool4 = MaxPooling2D(pool_size=[2, 2], strides=None, padding='valid')(conv6)
# maxpool4 = Dropout(0.1)(maxpool4)
reshape = Reshape([100 , 6400])(maxpool3)
dense1 = Dense(units=512 , activation='relu' , use_bias=True , kernel_initializer='he_normal' , kernel_regularizer=regularizers.l2(1e-5))(reshape)
dense1 = BatchNormalization(epsilon=0.0002)(dense1)
dense1 = Dropout(0.2)(dense1)
#融入了attention机制于全连接层；
# dense1 = Dense(units=512, activation='relu', use_bias=True, kernel_initializer='he_normal')(reshape)
# attention_prob = Dense(units=512, activation='softmax', name='attention_vec')(dense1)
# attention_mul = multiply([dense1, attention_prob])
#
# dense1 = BatchNormalization(epsilon=0.0002)(attention_mul)
# dense1 = Dropout(0.3)(dense1)
# dense2 = Dense(units=1024 , activation='relu' , use_bias=True , kernel_initializer='he_normal')(reshape)
# dense2 = BatchNormalization(epsilon=0.0002)(dense2)
# dense2 = Dropout(0.2)(dense2)
dense2 = Dense(units=self.MS_OUTPUT_SIZE , use_bias=True , kernel_initializer='he_normal' , kernel_regularizer=regularizers.l2(1e-5))(dense1)
y_pred = Activation(activation='softmax' , name='activation')(dense2)
model_data = Model(inputs=input_data , outputs=y_pred)
# model_data.summary()
labels = Input(shape=[self.label_max_string_length], name='labels', dtype='float32')
input_length = Input(shape=[1], name='input_length', dtype='int64')
label_length = Input(shape=[1], name='label_length', dtype='int64')
loss_out = Lambda(self.ctc_lambda_func, output_shape=[1, ], name='ctc')([y_pred , labels, input_length, label_length])
model = Model(inputs=[input_data, labels, input_length, label_length], outputs=loss_out)
# model.summary()
sgd = SGD(lr=0.0005, decay=1e-6, momentum=0.9, nesterov=True, clipnorm=5)
ada_d = Adadelta(lr=0.0005 , rho=0.95, epsilon=1e-6)
adam = Adam(lr=0.0001 , epsilon=1e-6)
model.compile(optimizer=adam , loss={'ctc': lambda y_true, y_pred: y_pred})
print('==========================模型创建成功=================================')
return model, model_data
def ctc_lambda_func(self , args):
y_pred , labels , input_length , label_length = args
y_pred = y_pred[: , : , :]
return K.ctc_batch_cost(y_true=labels , y_pred=y_pred , input_length=input_length , label_length=label_length)
def train_model(self , datapath , epoch=4 , save_step=2000 , batch_size=8):
data = DataSpeech(datapath , 'train')
num_data = data.get_datanum()
yielddatas = data.data_generator(batch_size , self.AUDIO_LENGTH)
for epoch in range(epoch):
print('[*running] train epoch %d .' % epoch)
n_step = 0
while True:
try:
print('[*message] epoch %d , Having training data %d+' % (epoch , n_step * save_step))
self._model.fit_generator(yielddatas , save_step)
n_step += 1
except StopIteration:
print('======================Error StopIteration==============================')
break
self.save_model(comments='_e_' + str(epoch) + '_step_' + str(n_step * save_step))
self.test_model(datapath=self.datapath , str_dataset='train' , data_count=4)
self.test_model(datapath=self.datapath , str_dataset='dev' , data_count=16)
def load_model(self , filename='model_speech_e_0_step_16000.model'):
self._model.load_weights(filename)
self.base_model.load_weights(filename + '.base')
def test_model(self , datapath='' , str_dataset='dev' , data_count=1):
data = DataSpeech(self.datapath , str_dataset)
num_data = data.get_datanum()
# print num_data
if data_count <=0 and data_count > num_data:
data_count = num_data
try:
ran_num = random.randint(0 , num_data - 1)
words_num = 0.
word_error_num = 0.
for i in range(data_count):
data_input , data_labels = data.get_data((ran_num + i) % num_data)
# print data_input
num_bias = 0
while data_input.shape[0] > self.AUDIO_LENGTH:
print('[*Error] data input is too long %d' % ((ran_num + i) % num_data))
num_bias += 1
data_input , data_labels = data.get_data((ran_num + i + num_bias) % num_data)
pre = self.predict(data_input=data_input , input_len=data_input.shape[0] // 8)
words_n = data_labels.shape[0]
words_num += words_n
edit_distance = get_edit_distance(data_labels , pre)
if edit_distance <= words_n:
word_error_num += edit_distance
else:
word_error_num += words_n
# print type(words_num)
print('[*Test Result] Speech Recognition ' + str_dataset + ' set word error ratio : ' + str(word_error_num / words_num * 100) , '%')
except StopIteration:
print('=======================Error StopIteration 01======================')
def save_model(self , filename='/home/zhangwei/speech_model/speech_model' , comments=''):
self._model.save_weights(filename + comments + '.model')
self.base_model.save_weights(filename + comments + '.model.base')
f = open('steps24.txt' , 'w')
f.write(filename + comments)
f.close()
def predict(self , data_input , input_len):
batch_size = 1
in_len = np.zeros((batch_size) , dtype=np.int32)
in_len[0] = input_len
x_in = np.zeros(shape=[batch_size , 1600 , self.AUDIO_FEATURE_LENGTH , 1] , dtype=np.float)
for i in range(batch_size):
x_in[i , 0 : len(data_input)] = data_input
base_pred = self.base_model.predict(x=x_in)
base_pred = base_pred[: , : , :]
r = K.ctc_decode(base_pred , in_len , greedy=True , beam_width=100 , top_paths=1)
r1 = K.get_value(r[0][0])
r1 = r1[0]
return r1
def redognize_speech(self , wavsignal , fs):
data_input = get_frequency_feature(wavsignal , fs)
input_length = len(data_input)
input_length = input_length // 8
data_input = np.array(data_input , dtype=np.float)
data_input = data_input.reshape(data_input.shape[0] , data_input.shape[1] , 1)
r1 = self.predict(data_input , input_length)
# print r1
list_symbol_dic = get_list_symbol(self.datapath)
r_str = []
for i in r1:
r_str.append(list_symbol_dic[i])
return r_str
def recognize_speech_fromfile(self , filename):
wavsignal , fs = read_wav_data(filename)
r = self.redognize_speech(wavsignal , fs)
return r
if __name__ == '__main__':
import tensorflow as tf
from keras.backend.tensorflow_backend import set_session
config = tf.ConfigProto()
config.gpu_options.per_process_gpu_memory_fraction = 0.1
set_session(tf.Session(config=config))
datapath = '/home/zhangwei/PycharmProjects/ASR_Thchs30/data_list/'
speech = ModelSpeech(datapath=datapath)
# speech.creat_model()
speech.train_model(datapath=datapath)