自定义损失函数
'''
return K.mean(K.binary_crossentropy(y_pred, y_true), axis=-1)
def my_loss2(y_true, y_pred):
'''
自定义损失函数
'''
cost = -K.mean(y_true * K.log(y_pred) + (1 - y_true) * K.log(1 - y_pred))
return cost
if __name__ == '__main__':
train_X_org, train_Y_org, test_X_org, test_Y_org, classes = kt_utils.load_dataset(
)
# step1 对数据进行归一化处理
train_X = train_X_org / 255
test_X = test_X_org / 255
# step2 对标签数据进行转置处理，在keras中采用(样本数， 标签)这样的输入！！！
train_Y = train_Y_org.T
test_Y = test_Y_org.T
# step3 创建一个模型
input_shape = train_X.shape[1:] # 64 X 64 X 3
happy_model = HappyModel(input_shape)
# step4 编译模型（定义训练方式）
# happy_model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])
使用keras来构建一个CNN模型
"""
from keras.layers import Input, Dense, Activation, ZeroPadding2D, BatchNormalization, Flatten, Conv2D
from keras.layers import MaxPooling2D
from keras.models import Model
import keras.backend as K
import os
import time
try:
from kt_utils import load_dataset
except ImportError:
raise ImportError('The file is not found. Please check the file name!')
K.set_image_data_format('channels_last')
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
X_train_orig, Y_train_orig, X_test_orig, Y_test_orig, classes = load_dataset()
# Normalize image vectors
X_train = X_train_orig / 255.
X_test = X_test_orig / 255.
# Reshape
Y_train = Y_train_orig.T
Y_test = Y_test_orig.T
print("number of training examples = " + str(X_train.shape[0]))
print("number of test examples = " + str(X_test.shape[0]))
print("X_train shape: " + str(X_train.shape))
print("Y_train shape: " + str(Y_train.shape))
print("X_test shape: " + str(X_test.shape))
print("Y_test shape: " + str(Y_test.shape))
X = BatchNormalization(axis=3,name=bn_name_base+"2c")(X)
#捷径
X_shortcut = Conv2D(filters=F3, kernel_size=(1,1), strides=(s,s), padding="valid",
name=conv_name_base+"1", kernel_initializer=glorot_uniform(seed=0))(X_shortcut)
X_shortcut = BatchNormalization(axis=3,name=bn_name_base+"1")(X_shortcut)
#最后一步
X = Add()([X,X_shortcut])
X = Activation("relu")(X)
return X
if __name__ == '__main__':
X_train_orig, Y_train_orig, X_test_orig, Y_test_orig, classes = kt_utils.load_dataset()
# Normalize image vectors
X_train = X_train_orig/255.
X_test = X_test_orig/255.
# Reshape
Y_train = Y_train_orig.T
Y_test = Y_test_orig.T
index = 3
plt.imshow(X_train_orig[index])
plt.show()
"""
#创建一个模型实体
happy_model = HappyModel(X_train.shape[1:])
#编译模型
