def train_with_augmentation():
datagen = ImageDataGenerator(rescale=1. / 255,
rotation_range=30.,
horizontal_flip=True)
model = vgg13()
(X_train, Y_train), (X_test, Y_test), (X_validation,
Y_validation) = fer2013()
if load_weights:
model.load_weights('model_vgg_13_aug.h5')
history = model.fit_generator(datagen.flow(X_train,
Y_train,
batch_size=batch_size),
samples_per_epoch=50000,
nb_epoch=nb_epoch,
validation_data=(X_validation, Y_validation))
predictions = model.predict(X_test, batch_size=batch_size, verbose=1)
historic(history)
confusion_matrix(predictions, Y_test)
if save_weights:
model.save_weights('model_vgg_13_aug.h5')
def train_without_augmentation_ck():
model = vgg16()
(X_train, Y_train), (X_test, Y_test), (X_validation, Y_validation) = ck()
if load_weights:
model.load_weights('model_vgg_16_ck.h5')
history = model.fit(X_train,
Y_train,
batch_size=batch_size,
nb_epoch=nb_epoch,
validation_data=(X_validation, Y_validation),
shuffle=True)
predictions = model.predict(X_test, batch_size=batch_size, verbose=1)
evaluates = model.evaluate(X_test, Y_test)
historic(history)
confusion_matrix(predictions, Y_test)
if save_weights:
model.save_weights('model_vgg_16_ck.h5')
def train_without_augmentation():
model = vgg16(lr=0.0001, dropout_in=0.25, dropout_out=0.5)
(X_train, Y_train), (X_test, Y_test), (X_validation,
Y_validation) = fer2013()
if load_weights:
model.load_weights('model_vgg_16_eq.h5')
history = model.fit(X_train,
Y_train,
batch_size=batch_size,
nb_epoch=nb_epoch,
validation_data=(X_validation, Y_validation),
shuffle=True)
if save_weights:
model.save_weights('vgg16_fer2013_np.h5')
predictions = model.predict(X_test, batch_size=batch_size, verbose=1)
evaluates = model.evaluate(X_test, Y_test)
historic(history)
confusion_matrix(predictions, Y_test)
def train_with_augmentation_ck():
datagen = ImageDataGenerator(rotation_range=10., horizontal_flip=True)
model = vgg16(lr=0.00005)
(X_train, Y_train), (X_test, Y_test), (X_validation, Y_validation) = ck()
if load_weights:
model.load_weights('model_vgg_16_eq.h5')
history = model.fit_generator(datagen.flow(X_train,
Y_train,
batch_size=batch_size),
samples_per_epoch=2000,
nb_epoch=nb_epoch,
validation_data=(X_validation, Y_validation))
predictions = model.predict(X_test, batch_size=batch_size, verbose=1)
historic(history)
confusion_matrix(predictions, Y_test)
if save_weights:
model.save_weights('model_vgg_16_ft_ck.h5')
def test(model=None):
if model == None:
model = basic()
model.load_weights('basic_fer2013.h5')
(X_train, Y_train), (X_test, Y_test), (X_validation, Y_validation) = ck()
scores = model.evaluate(X_train, Y_train)
print(scores)
predictions = model.predict(X_train, batch_size=batch_size, verbose=1)
confusion_matrix(predictions, Y_train)
scores = model.evaluate(X_validation, Y_validation)
print(scores)
predictions = model.predict(X_validation, batch_size=batch_size, verbose=1)
confusion_matrix(predictions, Y_validation)
scores = model.evaluate(X_test, Y_test)
print(scores)
predictions = model.predict(X_test, batch_size=batch_size, verbose=1)
confusion_matrix(predictions, Y_test)
(X_train, Y_train), (X_test, Y_test), (X_validation,
Y_validation) = fer2013()
scores = model.evaluate(X_train, Y_train)
print(scores)
predictions = model.predict(X_train, batch_size=batch_size, verbose=1)
confusion_matrix(predictions, Y_train)
scores = model.evaluate(X_validation, Y_validation)
print(scores)
predictions = model.predict(X_validation, batch_size=batch_size, verbose=1)
confusion_matrix(predictions, Y_validation)
scores = model.evaluate(X_test, Y_test)
print(scores)
predictions = model.predict(X_test, batch_size=batch_size, verbose=1)
confusion_matrix(predictions, Y_test)