douda_slim_gpu.py

#-*-coding:utf-8 -*-
import tensorflow as tf  # 0.12
import numpy as np
import os
from collections import Counter
import librosa  # https://github.com/librosa/librosa
 
# 训练样本路径
wav_path = 'data/wav/train'
label_file = 'data/doc/trans/train.word.txt'
 
# 获得训练用的wav文件路径列表
def get_wav_files(wav_path=wav_path):
	wav_files = []
	for (dirpath, dirnames, filenames) in os.walk(wav_path):
		for filename in filenames:
			if filename.endswith('.wav') or filename.endswith('.WAV'):
				filename_path = os.sep.join([dirpath, filename])
				if os.stat(filename_path).st_size < 240000:  # 剔除掉一些小文件
					continue
				wav_files.append(filename_path)
	return wav_files
 
wav_files = get_wav_files()
 
# 读取wav文件对应的label
def get_wav_lable(wav_files=wav_files, label_file=label_file):
	labels_dict = {}
	with open(label_file, 'r') as f:
		for label in f:
			label = label.strip('\n')
			label_id = label.split(' ', 1)[0]
			label_text = label.split(' ', 1)[1]
			labels_dict[label_id] = label_text
 
	labels = []
	new_wav_files = []
	for wav_file in wav_files:
		wav_id = os.path.basename(wav_file).split('.')[0]
		if wav_id in labels_dict:
			labels.append(labels_dict[wav_id])
			new_wav_files.append(wav_file)
 
	return new_wav_files, labels
 
wav_files, labels = get_wav_lable()
print("样本数:", len(wav_files))  # 8911
#print(wav_files[0], labels[0])
# wav/train/A11/A11_0.WAV -> 绿 是 阳春 烟 景 大块 文章 的 底色 四月 的 林 峦 更是 绿 得 鲜活 秀媚 诗意 盎然
 
# 词汇表(参看练习1和7)
all_words = []
for label in labels:
	all_words += [word for word in label]
counter = Counter(all_words)
count_pairs = sorted(counter.items(), key=lambda x: -x[1])
 
words, _ = zip(*count_pairs)
words_size = len(words)
print('词汇表大小:', words_size)
 
word_num_map = dict(zip(words, range(len(words))))
to_num = lambda word: word_num_map.get(word, len(words))
labels_vector = [ list(map(to_num, label)) for label in labels]
#print(wavs_file[0], labels_vector[0])
#wav/train/A11/A11_0.WAV -> [479, 0, 7, 0, 138, 268, 0, 222, 0, 714, 0, 23, 261, 0, 28, 1191, 0, 1, 0, 442, 199, 0, 72, 38, 0, 1, 0, 463, 0, 1184, 0, 269, 7, 0, 479, 0, 70, 0, 816, 254, 0, 675, 1707, 0, 1255, 136, 0, 2020, 91]
#print(words[479]) #绿
label_max_len = np.max([len(label) for label in labels_vector])
print('最长句子的字数:', label_max_len)
 
wav_max_len = 673
#for wav in wav_files:
#    wav, sr = librosa.load(wav, mono=True)
#    mfcc = np.transpose(librosa.feature.mfcc(wav, sr), [1,0])
#    if len(mfcc) > wav_max_len:
#        wav_max_len = len(mfcc)
print("最长的语音:", wav_max_len)
 
batch_size = 16
n_batch = len(wav_files) // batch_size
 
# 获得一个batch
pointer = 0
def get_next_batches(batch_size):
    global pointer
    batches_wavs = []
    batches_labels = []
    for i in range(batch_size):
        wav, sr = librosa.load(wav_files[pointer], mono=True)
        mfcc = np.transpose(librosa.feature.mfcc(wav, sr), [1,0])
        batches_wavs.append(mfcc.tolist())
        batches_labels.append(labels_vector[pointer])
        pointer += 1
 
    # 补零对齐
    for mfcc in batches_wavs:
        while len(mfcc) < wav_max_len:
            mfcc.append([0]*20)
    for label in batches_labels:
        while len(label) < label_max_len:
            label.append(0)
    return batches_wavs, batches_labels
 
X = tf.placeholder(dtype=tf.float32, shape=[batch_size, None, 20])
sequence_len = tf.reduce_sum(tf.cast(tf.not_equal(tf.reduce_sum(X, reduction_indices=2), 0.), tf.int32), reduction_indices=1)
Y = tf.placeholder(dtype=tf.int32, shape=[batch_size, None])
 
# conv1d_layer
conv1d_index = 0
def conv1d_layer(input_tensor, size, dim, activation, scale, bias):
    global conv1d_index
    with tf.variable_scope('conv1d_' + str(conv1d_index)):
        W = tf.get_variable('W', (size, input_tensor.get_shape().as_list()[-1], dim), dtype=tf.float32, initializer=tf.random_uniform_initializer(minval=-scale, maxval=scale))
        if bias:
            b = tf.get_variable('b', [dim], dtype=tf.float32, initializer=tf.constant_initializer(0))
        out = tf.nn.conv1d(input_tensor, W, stride=1, padding='SAME') + (b if bias else 0)
        if not bias:
            beta = tf.get_variable('beta', dim, dtype=tf.float32, initializer=tf.constant_initializer(0))
            gamma = tf.get_variable('gamma', dim, dtype=tf.float32, initializer=tf.constant_initializer(1))
            mean_running = tf.get_variable('mean', dim, dtype=tf.float32, initializer=tf.constant_initializer(0))
            variance_running = tf.get_variable('variance', dim, dtype=tf.float32, initializer=tf.constant_initializer(1))
            mean, variance = tf.nn.moments(out, axes=range(len(out.get_shape()) - 1))
            def update_running_stat():
                decay = 0.99
                update_op = [mean_running.assign(mean_running * decay + mean * (1 - decay)), variance_running.assign(variance_running * decay + variance * (1 - decay))]
                with tf.control_dependencies(update_op):
                    return tf.identity(mean), tf.identity(variance)
                m, v = tf.cond(tf.Variable(False, trainable=False, collections=[tf.GraphKeys.LOCAL_VARIABLES]), update_running_stat, lambda: (mean_running, variance_running))
                out = tf.nn.batch_normalization(out, m, v, beta, gamma, 1e-8)
        if activation == 'tanh':
            out = tf.nn.tanh(out)
        if activation == 'sigmoid':
            out = tf.nn.sigmoid(out)
 
        conv1d_index += 1
        return out
# aconv1d_layer
aconv1d_index = 0
def aconv1d_layer(input_tensor, size, rate, activation, scale, bias):
    global aconv1d_index
    with tf.variable_scope('aconv1d_' + str(aconv1d_index)):
        shape = input_tensor.get_shape().as_list()
        W = tf.get_variable('W', (1, size, shape[-1], shape[-1]), dtype=tf.float32, initializer=tf.random_uniform_initializer(minval=-scale, maxval=scale))
        if bias:
            b = tf.get_variable('b', [shape[-1]], dtype=tf.float32, initializer=tf.constant_initializer(0))
        out = tf.nn.atrous_conv2d(tf.expand_dims(input_tensor, dim=1), W, rate=rate, padding='SAME')
        out = tf.squeeze(out, [1])
        if not bias:
            beta = tf.get_variable('beta', shape[-1], dtype=tf.float32, initializer=tf.constant_initializer(0))
            gamma = tf.get_variable('gamma', shape[-1], dtype=tf.float32, initializer=tf.constant_initializer(1))
            mean_running = tf.get_variable('mean', shape[-1], dtype=tf.float32, initializer=tf.constant_initializer(0))
            variance_running = tf.get_variable('variance', shape[-1], dtype=tf.float32, initializer=tf.constant_initializer(1))
            mean, variance = tf.nn.moments(out, axes=range(len(out.get_shape()) - 1))
            def update_running_stat():
                decay = 0.99
                update_op = [mean_running.assign(mean_running * decay + mean * (1 - decay)), variance_running.assign(variance_running * decay + variance * (1 - decay))]
                with tf.control_dependencies(update_op):
                    return tf.identity(mean), tf.identity(variance)
                m, v = tf.cond(tf.Variable(False, trainable=False, collections=[tf.GraphKeys.LOCAL_VARIABLES]), update_running_stat, lambda: (mean_running, variance_running))
                out = tf.nn.batch_normalization(out, m, v, beta, gamma, 1e-8)
        if activation == 'tanh':
            out = tf.nn.tanh(out)
        if activation == 'sigmoid':
            out = tf.nn.sigmoid(out)
 
        aconv1d_index += 1
        return out
# 定义神经网络
def speech_to_text_network(n_dim=128, n_blocks=3):
    out = conv1d_layer(input_tensor=X, size=1, dim=n_dim, activation='tanh', scale=0.14, bias=False)
    # skip connections
    def residual_block(input_sensor, size, rate):
            conv_filter = aconv1d_layer(input_sensor, size=size, rate=rate, activation='tanh', scale=0.03, bias=False)
            conv_gate = aconv1d_layer(input_sensor, size=size, rate=rate,  activation='sigmoid', scale=0.03, bias=False)
            out = conv_filter * conv_gate
            out = conv1d_layer(out, size=1, dim=n_dim, activation='tanh', scale=0.08, bias=False)
            return out + input_sensor, out
    skip = 0
    for _ in range(n_blocks):
        for r in [1, 2, 4, 8, 16]:
            out, s = residual_block(out, size=7, rate=r)
            skip += s
 
    logit = conv1d_layer(skip, size=1, dim=skip.get_shape().as_list()[-1], activation='tanh', scale=0.08, bias=False)
    logit = conv1d_layer(logit, size=1, dim=words_size, activation=None, scale=0.04, bias=True)
 
    return logit
 
class MaxPropOptimizer(tf.train.Optimizer):
    def __init__(self, learning_rate=0.001, beta2=0.999, use_locking=False, name="MaxProp"):
        super(MaxPropOptimizer, self).__init__(use_locking, name)
        self._lr = learning_rate
        self._beta2 = beta2
        self._lr_t = None
        self._beta2_t = None
    def _prepare(self):
        self._lr_t = tf.convert_to_tensor(self._lr, name="learning_rate")
        self._beta2_t = tf.convert_to_tensor(self._beta2, name="beta2")
    def _create_slots(self, var_list):
        for v in var_list:
            self._zeros_slot(v, "m", self._name)
    def _apply_dense(self, grad, var):
        lr_t = tf.cast(self._lr_t, var.dtype.base_dtype)
        beta2_t = tf.cast(self._beta2_t, var.dtype.base_dtype)
        if var.dtype.base_dtype == tf.float16:
            eps = 1e-7
        else:
            eps = 1e-8
        m = self.get_slot(var, "m")
        m_t = m.assign(tf.maximum(beta2_t * m + eps, tf.abs(grad)))
        g_t = grad / m_t
        var_update = tf.assign_sub(var, lr_t * g_t)
        return tf.group(*[var_update, m_t])
    def _apply_sparse(self, grad, var):
        return self._apply_dense(grad, var)
 
def train_speech_to_text_network():
    logit = speech_to_text_network()
 
    # CTC loss
    indices = tf.where(tf.not_equal(tf.cast(Y, tf.float32), 0.))
    target = tf.SparseTensor(indices=indices, values=tf.gather_nd(Y, indices) - 1, shape=tf.cast(tf.shape(Y), tf.int64))
    loss = tf.nn.ctc_loss(logit, target, sequence_len, time_major=False)
    # optimizer
    lr = tf.Variable(0.001, dtype=tf.float32, trainable=False)
    optimizer = MaxPropOptimizer(learning_rate=lr, beta2=0.99)
    var_list = [t for t in tf.trainable_variables()]
    gradient = optimizer.compute_gradients(loss, var_list=var_list)
    optimizer_op = optimizer.apply_gradients(gradient)
 
    with tf.Session(config=tf.ConfigProto(device_count={'gpu':0})) as sess:
        sess.run(tf.global_variables_initializer())
 
        saver = tf.train.Saver(tf.global_variables())
 
        for epoch in range(16):
            sess.run(tf.assign(lr, 0.001 * (0.97 ** epoch)))
 
            global pointer
            pointer = 0
            for batch in range(n_batch):
                batches_wavs, batches_labels = get_next_batches(batch_size)
                train_loss, _ = sess.run([loss, optimizer_op], feed_dict={X: batches_wavs, Y: batches_labels})
                print(epoch, batch, train_loss)
            if epoch % 5 == 0:
                saver.save(sess, 'speech.module', global_step=epoch)
 
# 训练
train_speech_to_text_network()
 
# 语音识别
# 把batch_size改为1
def speech_to_text(wav_file):
    wav, sr = librosa.load(wav_file, mono=True)
    mfcc = np.transpose(np.expand_dims(librosa.feature.mfcc(wav, sr), axis=0), [0,2,1])
 
    logit = speech_to_text_network()
 
    saver = tf.train.Saver()
    with tf.Session(config=tf.ConfigProto(device_count={'gpu':0})) as sess:
        saver.restore(sess, tf.train.latest_checkpoint('.'))
 
        decoded = tf.transpose(logit, perm=[1, 0, 2])
        decoded, _ = tf.nn.ctc_beam_search_decoder(decoded, sequence_len, merge_repeated=False)
        predict = tf.sparse_to_dense(decoded[0].indices, decoded[0].shape, decoded[0].values) + 1
        output = sess.run(decoded, feed_dict={X: mfcc})        
        print(output)