smelloffear-deeplearning/labels_audio.py at master · TomFevrier/smelloffear-deeplearning · GitHub

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
def warn(*args, **kwargs):
	pass
import warnings
warnings.warn = warn

import keras
from keras.models import Sequential, load_model
from keras.layers import Dense
from keras.callbacks import Callback
from sklearn.metrics import f1_score, precision_score, recall_score, classification_report
import numpy as np
import sys, os, argparse
import math
import h5py

os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'

parser = argparse.ArgumentParser()
parser.add_argument('model_version',
					help="Model version")
parser.add_argument('-b', '--batch_size',
					help="Batch size",
					type=int,
					default=64)
parser.add_argument('-e', '--epochs',
					help="Number of epochs",
					type=int,
					default=100)
args = parser.parse_args()

batch_size = args.batch_size
epochs = args.epochs

model_name = "./models/labels_audio_v" + args.model_version

input_dim = 193
nb_classes = 13

class ClassificationReport(Callback):

	def __init__(self, test_data):
		self.test_data = test_data

	def on_epoch_end(self, epoch, logs={}):
		predict = np.asarray(self.model.predict(self.test_data[0]).round())
		target = self.test_data[1]
		f1 = f1_score(target, predict, average='weighted')
		recall = recall_score(target, predict, average='weighted')
		precision = precision_score(target, predict, average='weighted')
		metrics = "-- f1: " + "{0:.3f}".format(f1) + " -- precision: " + "{0:.3f}".format(precision) + " -- recall " + "{0:.3f}".format(recall)
		print(metrics)

		path = "./csv/performance_labels_audio_v" + args.model_version + ".csv"
		if not os.path.exists(path):
			fd = open(path, 'w+')
			fd.write("f1,precision,recall\n")
			fd.close()
		row = str(round(f1, 4)) + "," + str(round(precision, 4)) + "," + str(round(recall, 4)) + "\n"
		fd = open(path, 'a')
		fd.write(row)
		fd.close()


def initiate_model():
	if not os.path.exists(model_name):
		model = Sequential()
		model.add(Dense(280, input_dim=input_dim))
		model.add(Dense(300, activation='tanh'))
		model.add(Dense(nb_classes, activation='sigmoid'))

		model.compile(loss=keras.losses.binary_crossentropy,
					  optimizer=keras.optimizers.Adadelta())

	else:
		model = load_model(model_name)

	model.summary()
	print()
	return model


def create_datasets():
	films = ['Buddy', 'Hobbit', 'Machete', 'Mitty', 'Paranormal', 'Tribute']

	for film in films:
		features = np.load("./data/audio/" + film + ".npy")

		useful_classes = (0, 4, 14, 15, 16, 17, 18, 21, 23, 24, 28, 33, 41)
		labels = np.genfromtxt("./data/labels/" + film + "_transposed.csv", delimiter=',', usecols=useful_classes)

		limit = min(features.shape[0], labels.shape[0])
		limit_training_set = math.floor(0.8 * limit)

		if film == films[0]:
			x_train = features[:limit_training_set]
			y_train = labels[:limit_training_set]
			x_test = features[limit_training_set:limit]
			y_test = labels[limit_training_set:limit]
		else:
			x_train = np.concatenate((x_train, features[:limit_training_set]))
			y_train = np.concatenate((y_train, labels[:limit_training_set]))
			x_test = np.concatenate((x_test, features[limit_training_set:limit]))
			y_test = np.concatenate((y_test, labels[limit_training_set:limit]))

	return x_train, y_train, x_test, y_test


def evaluate_model(y_test, y_pred):
	target_names = ['suspense', 'comedy', 'drama', 'everyday life', 'dream', 'landscape', 'conversation', 'conversation main actor', 'action', 'death', 'running', 'blood (violence)', 'sudden shock']

	print(classification_report(y_test, y_pred, target_names=target_names))


def main():
	model = initiate_model()

	x_train, y_train, x_test, y_test = create_datasets()

	x_train = x_train.astype('float32')
	x_test = x_test.astype('float32')

	print('x_train shape:', x_train.shape)
	print(x_train.shape[0], 'train samples')
	print(x_test.shape[0], 'test samples')
	print()

	report = ClassificationReport((x_test, y_test))

	model.fit(x_train, y_train,
	          batch_size=batch_size,
	          epochs=epochs,
	          verbose=1,
	          validation_data=(x_test, y_test),
			  callbacks=[report])

	y_pred = np.asarray(model.predict(x_test).round())
	evaluate_model(y_test, y_pred)

	model.save(model_name)


main()