try:
if not is_light_net:
face = cv2.resize(rgb_face,(target_size),interpolation=cv2.INTER_CUBIC)
face = preprocess_input(face, False)
face = np.expand_dims(face, 0)
else:
face = cv2.resize(gray_face, (target_size),interpolation=cv2.INTER_CUBIC)
face = preprocess_input(face, False)
face = np.expand_dims(face, 0)
face = np.expand_dims(face, -1)
except:
continue
start_time = time.clock()
if is_single_task:
emotion_label = emotion_model.predict(face)
gender_label = gender_model.predict(face)
pose_label = pose_model.predict(face)
age_label= age_model.predict(face)
else:
if is_EGA:
emotion_label,gender_label,age_label = EGA_multi_model.predict(face)
elif is_EPGA:
emotion_label,pose_label,gender_label,age_label = EPGA_multi_model.predict(face)
print('predicted result')
else:
emotion_label,pose_label,age_label=EPA_multi_model.predict(face)
end_time = time.clock()
print('spend_time:',end_time-start_time)
if is_single_task:
interpolation=cv2.INTER_CUBIC)
face = preprocess_input(face, False)
face = np.expand_dims(face, 0)
face = np.expand_dims(face, -1)
except:
continue
start_time = time.clock()
if is_single_task:
emotion_label = emotion_model.predict(face)
gender_label = gender_model.predict(face)
pose_label = pose_model.predict(face)
age_label = age_model.predict(face)
else:
if is_EGA:
emotion_label, gender_label, age_label = EGA_multi_model.predict(
face)
elif is_EPA:
emotion_label, pose_label, age_label = EPA_multi_model.predict(
face)
elif is_EPGA:
print('here')
emotion_label, pose_label, gender_label, age_label = EPGA_multi_model.predict(
face)
end_time = time.clock()
print('spend_time:', end_time - start_time)
if is_single_task:
emotion_label_arg = np.argmax(emotion_label)
gender_label_arg = np.argmax(gender_label)
age_label_arg = np.argmax(age_label)
pose_label_arg = np.argmax(pose_label)
}
if __name__ == '__main__':
tf.keras.layers.Conv2D()
data_x, data_y = loadData(img_path)
#Classification
x_train, y_train, x_val, y_val, x_test, y_test = prePrec(data_x, data_y)
model_clas = Net('classification')
model_clas = train(model_clas,
x_train,
y_train,
x_val,
y_val,
parameters=parameter_set['classification'])[0]
print(model_clas.predict(x_test))
print('Accuracy on testset:{}'.format(
accuracy_score(y_test, model_clas.predict(x_test))))
#Regression
model_reg = Net('regression')
model_reg = train(model_reg,
x_train,
y_train,
x_val,
y_val,
parameters=parameter_set['regression'])[0]
mse = tf.keras.losses.MeanSquaredError()
loss = mse((y_val - 15) / 15., model_reg.predict(x_val).reshape(-1, ))
print('MSE on testset:{}'.format(loss))
#######################MobileNet-v2############################
