在转移学习:InceptionV3

时间:2019-02-15 10:51:47

标签: tensorflow keras deep-learning computer-vision transfer-learning

我正在学习转学。我的用例是对图像进行两类分类。我使用InceptionV3对图像进行分类。训练模型时,在每个时期我都会得到 nan 作为损失,而得到 0.0000e + 00 。我使用20个纪元是因为我的数据量很小:我得到了1000张训练图像和100张测试图像,每批5条记录。

from keras.applications.inception_v3 import InceptionV3
from keras.preprocessing import image
from keras.models import Model
from keras.layers import Dense, GlobalAveragePooling2D
from keras import backend as K

# create the base pre-trained model
base_model = InceptionV3(weights='imagenet', include_top=False)

# add a global spatial average pooling layer
x = base_model.output
x = GlobalAveragePooling2D()(x)
# let's add a fully-connected layer
x = Dense(1024, activation='relu')(x)

x = Dense(512, activation='relu')(x)
x = Dense(32, activation='relu')(x)
# and a logistic layer -- we have 2 classes
predictions = Dense(1, activation='softmax')(x)

# this is the model we will train
model = Model(inputs=base_model.input, outputs=predictions)


for layer in base_model.layers:
    layer.trainable = False

# we chose to train the top 2 inception blocks, i.e. we will freeze
# the first 249 layers and unfreeze the rest:
for layer in model.layers[:249]:
   layer.trainable = False
for layer in model.layers[249:]:
   layer.trainable = True

model.compile(loss="binary_crossentropy", optimizer="adam", metrics=["accuracy"])
from keras.preprocessing.image import ImageDataGenerator
train_datagen = ImageDataGenerator(
        rescale=1./255,
        shear_range=0.2,
        zoom_range=0.2,
        horizontal_flip=True)

test_datagen = ImageDataGenerator(rescale=1./255)

training_set = train_datagen.flow_from_directory(
        'C:/Users/Desktop/Transfer/train/',
        target_size=(64, 64),
        batch_size=5,
        class_mode='binary')

test_set = test_datagen.flow_from_directory(
        'C:/Users/Desktop/Transfer/test/',
        target_size=(64, 64),
        batch_size=5,
        class_mode='binary')

model.fit_generator(
        training_set,
        steps_per_epoch=1000,
        epochs=20,
        validation_data=test_set,
        validation_steps=100)

2 个答案:

答案 0 :(得分:1)

听起来您的梯度正在爆炸。可能有几个原因:

  • 检查输入是否正确生成。例如,使用save_to_dir
  • flow_from_directory参数
  • 由于批次大小为5,因此将steps_per_epoch1000固定为1000/5=200
  • 使用sigmoid激活代替softmax
  • 设置较低的亚当学习率;为此,您需要像adam = Adam(0.0001)一样分别创建优化器并将其传递到model.compile(..., optimizer=adam)
  • 尝试用VGG16代替InceptionV3

当您尝试以上所有方法时,请告知我们。

答案 1 :(得分:1)

使用Softmax进行激活对于单班没有意义。您的输出值将始终由其自身归一化,因此等于1。softmax的目的是使这些值的总和为1。在单个值的情况下,您将得到它==1。我相信您有时会得到零作为预测值,导致零除和NaN损失值。

您应该通过以下方式将类数更改为2:

  • predictions = Dense(2, activation='softmax')(x)
  • class_mode='categorical'中的{li> flow_from_directory
  • loss="categorical_crossentropy"

或在最后一层使用S型激活功能。