def lstm_example():
to_flatten = False
x_train, x_test, y_train, y_test, num_labels = extract_data(
flatten=to_flatten)
y_train = np_utils.to_categorical(y_train)
y_test_train = np_utils.to_categorical(y_test)
#train the model
# print('Starting LSTM')
# model = LSTM(input_shape=x_train[0].shape,
# num_classes=num_labels)
# print('x shape', x_train[0].shape)
# print('num labels',num_labels)
#
# model.train(x_train, y_train, x_test, y_test_train, n_epochs=50)
# model.evaluate(x_test, y_test)
# model.save_model()
newmodel = LSTM(input_shape=x_train[0].shape, num_classes=num_labels)
newmodel.load_model(to_load="./LSTM_best_model.h5")
newmodel.train(x_train, y_train, x_test, y_test_train, n_epochs=0)
audio_file = "../dataset/union-interview.wav"
section_by_section_analysis(model=newmodel,
audio_file=audio_file,
to_flatten=to_flatten)
def ml_example():
to_flatten = True
x_train, x_test, y_train, y_test, _ = extract_data(flatten=to_flatten)
model = NN()
print('Starting', model.name)
model.train(x_train, y_train)
model.evaluate(x_test, y_test)
filename = '../dataset/Sad/09b03Ta.wav'
print('prediction', model.predict_one(
get_feature_vector_from_mfcc(filename, flatten=to_flatten)),
'Actual 3')
def ml_example():
to_flatten = True
x_train, x_test, y_train, y_test, _ = extract_data(flatten=to_flatten)
model = RF()
print('Starting', model.name)
model.train(x_train, y_train)
filename = 'rf_finalized_model.sav'
pickle.dump(model, open(filename, 'wb'))
model.evaluate(x_test, y_test)
filename = '../AudioData/test--anger.wav'
print(
'prediction',
model.predict_one(
get_feature_vector_from_mfcc(filename, flatten=to_flatten)),
'Actual: Anger')
def lstm_example():
to_flatten = False
x_train, x_test, y_train, y_test, num_labels = extract_data(
flatten=to_flatten)
y_train = np_utils.to_categorical(y_train)
y_test_train = np_utils.to_categorical(y_test)
print('Starting LSTM')
model = LSTM(input_shape=x_train[0].shape, num_classes=num_labels)
# model.train(x_train, y_train, x_test, y_test_train, n_epochs=50)
# model.evaluate(x_test, y_test)
filename = 'dataset/Sad/09b03Ta.wav'
print(
'prediction',
model.predict_one(
get_feature_vector_from_mfcc(filename, flatten=to_flatten)),
'Actual 3')
def lstm_saved_model():
to_flatten = False
x_train, x_test, y_train, y_test, num_labels = extract_data(
flatten=to_flatten)
y_train = np_utils.to_categorical(y_train)
y_test_train = np_utils.to_categorical(y_test)
print('Starting LSTM')
model = LSTM(input_shape=x_train[0].shape,
num_classes=num_labels, save_path="./lstm.h5")
print("Loading")
model.restore_model()
# filename = '../dataset/03a01Wa.wav'
filename = '../dataset/sad.wav'
print("\nPredicted: {}\nActual: {}".format(
get_class_name(
model.predict_one(get_feature_vector_from_mfcc(filename, flatten=to_flatten))),
get_class_name(3)))
def ml_example():
to_flatten = True
x_train, x_test, y_train, y_test, _ = extract_data(flatten=to_flatten)
model = NN()
print('Starting', model.name)
model.train(x_train, y_train)
model.evaluate(x_test, y_test)
filename = '../dataset/Sad/09b03Ta.wav'
i = model.predict_one(
get_feature_vector_from_mfcc(filename, flatten=to_flatten))
if i == 0:
print("Prediction: Angry")
elif i == 1:
print("Prediction: Happy")
elif i == 2:
print("Prediction: Neutral")
else:
print("Prediction: Sad")
def cnn_example():
to_flatten = False
x_train, x_test, y_train, y_test, num_labels = extract_data(
flatten=to_flatten)
y_train = np_utils.to_categorical(y_train)
y_test_train = np_utils.to_categorical(y_test)
in_shape = x_train[0].shape
x_train = x_train.reshape(x_train.shape[0], in_shape[0], in_shape[1], 1)
x_test = x_test.reshape(x_test.shape[0], in_shape[0], in_shape[1], 1)
model = CNN(input_shape=x_train[0].shape,
num_classes=num_labels)
model.train(x_train, y_train, x_test, y_test_train)
model.evaluate(x_test, y_test)
filename = '../dataset/Sad/09b03Ta.wav'
print('prediction', model.predict_one(
get_feature_vector_from_mfcc(filename, flatten=to_flatten)),
'Actual 3')
print('CNN Done')
def lstm_example():
to_flatten = False
x_train, x_test, y_train, y_test, num_labels = extract_data(
flatten=to_flatten)
y_train = np_utils.to_categorical(y_train)
y_test_train = np_utils.to_categorical(y_test)
print('Starting LSTM')
model = LSTM(input_shape=x_train[0].shape, num_classes=num_labels)
print('Model Object Created, Going to train Model')
model.train(x_train, y_train, x_test, y_test_train, n_epochs=50)
print('Model Object trained, Going to evaluate Model')
model.evaluate(x_test, y_test)
print('Model Object Evaludated, Going to import testing audio')
filename = './dataset/Sad/09b03Ta.wav'
print(
'prediction',
model.predict_one(
get_feature_vector_from_mfcc(filename, flatten=to_flatten)),
'Actual 3')
print('Finishing LSTM')
def lstm_example():
to_flatten = False
x_train, x_test, y_train, y_test, num_labels = extract_data(
flatten=to_flatten)
y_train = np_utils.to_categorical(y_train)
y_test_train = np_utils.to_categorical(y_test)
print('Starting LSTM')
model = LSTM(input_shape=x_train[0].shape,
num_classes=num_labels, save_path="./lstm.h5")
print("Training")
model.train(x_train, y_train, x_test, y_test_train, n_epochs=10)
print("Evaluating")
model.evaluate(x_test, y_test)
model.save_model()
filename = '../dataset/Sad/09b03Ta.wav'
# filename = './laugh.wav'
print("\nPredicted: {}\nActual: {}".format(
get_class_name(
model.predict_one(
get_feature_vector_from_mfcc(filename, flatten=to_flatten))),
get_class_name(2)))
import sys
sys.path.insert(1, '../main')
import pickle
import utility
from common import extract_data
from mlmodel import NN
from utility import get_feature_vector_from_mfcc
filename = 'finalized_model.sav'
loaded_model = pickle.load(open(filename, 'rb'))
x_train, x_test, y_train, y_test, _ = extract_data(flatten=True)
loaded_model.evaluate(x_test, y_test)
filename = "../AudioData/testcaleb.wav"
print(
loaded_model.predict_one(
get_feature_vector_from_mfcc(filename, flatten=True)), "actual:Angry")
required=True,
type=int)
parser.add_argument('-r',
'--LEARNING_RATE',
help='learning rate',
required=True,
type=float)
parser.add_argument('-a',
'--ALPHA',
help='regularization coefficient',
required=True,
type=float)
args = parser.parse_args()
commondir = args.scriptdir
sys.path.append(commondir)
from common.utils import *
from common.extract_data import *
# load data into Pandas dataframe
data_dir = args.datadir
if not os.path.exists(os.path.join(data_dir, 'energy.csv')):
extract_data(data_dir)
energy = load_data(data_dir)
out_file = os.path.join(args.outdir, 'output.txt')
run(energy, args.T, args.LATENT_DIM_1, args.LATENT_DIM_2, args.BATCH_SIZE,
args.LEARNING_RATE, args.ALPHA, out_file)
def perceptron_example():
"""Perceptron example demonstrating online learning, and also evaluation
separate from training."""
tf.logging.set_verbosity(FLAGS.verbosity)
train_raw, test_raw, classes = get_data(FLAGS.data_dir)
# Set the output dimension according to the number of classes
FLAGS.output_dim = len(classes)
print("\nSplitting the training and test data into two pieces...")
# Seeding necessary for reproducibility.
np.random.seed(FLAGS.np_seed)
# Shuffle data to make the distribution of classes roughly stratified after splitting.
train_raw = shuffle(train_raw)
test_raw = shuffle(test_raw)
train1_raw, train2_raw = np.split(train_raw, 2)
test1_raw, test2_raw = np.split(test_raw, 2)
print("First split:")
x_train1_sentences, y_train1, x_test1_sentences, y_test1 = extract_data(
train1_raw, test1_raw)
print("\nProcessing the vocabulary...")
tic()
x_train1, x_test1, _, _, vocab_processor, n_words = process_vocabulary(
x_train1_sentences, x_test1_sentences, FLAGS, reuse=False)
toc()
# Train the model on the first split.
tic()
run_experiment(x_train1, y_train1, x_test1, y_test1,
bag_of_words_perceptron_model, 'train_and_evaluate', FLAGS)
toc()
# Next we perform incremental training with the 2nd half of the split data.
print("\nSecond split extends the vocabulary.")
x_train2_sentences, y_train2, x_test2_sentences, y_test2 = extract_data(
train2_raw, test2_raw)
# Extend vocab_processor with the newly added training vocabulary, and save the vocabulary processor for later use.
tic()
x_train2, x_test2, _, _, vocab_processor, n_words = process_vocabulary(
x_train2_sentences,
x_test2_sentences,
FLAGS,
reuse=False,
vocabulary_processor=vocab_processor,
extend=True)
toc()
# Train the model on the second split.
tic()
run_experiment(x_train2, y_train2, x_test2, y_test2,
bag_of_words_perceptron_model, 'train_and_evaluate', FLAGS)
toc()
# We may be interested in the model performance on the training data
# (e.g. to evaluate removable bias).
print("\nEvaluation of the model performance on the training data.:")
run_experiment(None, None, x_train1, y_train1,
bag_of_words_perceptron_model, 'evaluate', FLAGS)