Sound_Classifier_Neural_Network¶
In [1]:
import glob
import os
import librosa
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline
plt.style.use('ggplot')
import keras
from keras import models
from keras.layers import Dense, Dropout, Activation, Flatten
from keras.layers import Convolution2D, MaxPooling2D
from keras.utils import to_categorical
keras.__version__
Out[1]:
In [2]:
spl = 0
parent_dir = '../data/UrbanSound8K/audio/'
if spl:
sub_dir = 'samples/'
else:
sub_dir = 'train/'
file_name = '*.wav'
files = glob.glob(os.path.join(parent_dir, sub_dir, file_name))
test_files = glob.glob(os.path.join(parent_dir, 'fold10/', file_name))
files[1]
Out[2]:
In [3]:
def parse_audio(files):
features, labels = np.empty((0,52)), np.empty(0)
for file in files:
# extract label from filename
labels = np.append(labels, file.split('\\')[1].split('-')[1])
# extract features from audio file
x, sr = librosa.load(file)
stft = np.abs(librosa.stft(x))
mfccs = np.mean(librosa.feature.mfcc(y=x, sr=sr, n_mfcc=40).T, axis=0)
chroma = np.mean(librosa.feature.chroma_stft(S=stft,sr=sr).T, axis=0)
feature = np.hstack((mfccs,chroma))
# print(chroma.shape)
# print(mfccs.shape)
# print(feature.shape)
features = np.vstack((features, feature))
return np.array(features), np.array(labels, dtype=np.int)
In [4]:
features, labels = parse_audio(files)
train_labels = to_categorical(labels)
print(features.shape)
print(labels.shape)
test_features, test_labels = parse_audio(test_files)
test_labels = to_categorical(test_labels)
In [22]:
np.savetxt('train_features',features)
np.savetxt('train_labels',train_labels)
np.savetxt('test_features',test_features)
np.savetxt('test_labels',test_labels)
Build and compile model¶
In [41]:
# model = models.Sequential()
# model.add(layers.Dense(256, activation='relu', input_shape=(52,)))
# model.add(layers.Dense(50, activation='relu'))
# # model.add(layers.Dense(512, activation='relu'))
# model.add(layers.Dense(10, activation='softmax'))
# model.compile(optimizer='Adagrad',
# loss='categorical_crossentropy',
# metrics=['accuracy'])
model = models.Sequential()
model.add(Dense(256,input_shape=(52,)))
model.add(Activation('relu'))
model.add(Dropout(0.5))
model.add(Dense(256))
model.add(Activation('relu'))
model.add(Dropout(0.5))
model.add(Dense(256))
model.add(Activation('relu'))
model.add(Dropout(0.5))
model.add(Dense(256))
model.add(Activation('relu'))
model.add(Dropout(0.5))
model.add(Dense(10))
model.add(Activation('softmax'))
model.compile(optimizer='Adam',
loss='categorical_crossentropy',
metrics=['accuracy'])
Train model¶
In [42]:
# from keras.callbacks import TensorBoard
# from keras.utils import plot_model
# tb = TensorBoard(log_dir='./logs', # log 目录
# histogram_freq=1, # 按照何等频率（epoch）来计算直方图，0为不计算
# batch_size=32, # 用多大量的数据计算直方图
# write_graph=True, # 是否存储网络结构图
# write_grads=False, # 是否可视化梯度直方图
# write_images=False,# 是否可视化参数
# embeddings_freq=0,
# embeddings_layer_names=None,
# embeddings_metadata=None)
# callbacks = [tb]
# history = model.fit(features, train_labels, epochs=50, batch_size=32, validation_data=(test_features, test_labels),callbacks=callbacks)
history = model.fit(features, train_labels, epochs=50, batch_size=32, validation_data=(test_features, test_labels))
In [48]:
print(history.history.keys())
fig = plt.figure(figsize=(20,5))
plt.subplot(121)
plt.plot(history.history['acc'])
plt.plot(history.history['val_acc'])
plt.title('model accuracy')
plt.ylabel('accuracy')
plt.xlabel('epoch')
plt.legend(['train', 'test'], loc='upper left')
plt.subplot(122)
plt.plot(history.history['loss'])
plt.plot(history.history['val_loss'])
plt.title('model loss')
plt.ylabel('loss')
plt.xlabel('epoch')
plt.legend(['train', 'test'], loc='lower left')
plt.show()
Test model¶
In [49]:
test_loss, test_acc = model.evaluate(test_features, test_labels)
test_acc
Out[49]: