Python 检查目标时出错:将FC层转换为Conv2D
我试图用卷积层替换VGG16网络末端的FC层。下面是我的代码:Python 检查目标时出错:将FC层转换为Conv2D,python,tensorflow,keras,deep-learning,conv-neural-network,Python,Tensorflow,Keras,Deep Learning,Conv Neural Network,我试图用卷积层替换VGG16网络末端的FC层。下面是我的代码: model2= Sequential() model2.add(Conv2D(4096, kernel_size=(8,8), activation="relu")) model2.add(Conv2D(4096, kernel_size=(1,1), activation="relu")) model2.add(Conv2D(16, kernel_size=(1,1), activation="softmax")) model
model2= Sequential()
model2.add(Conv2D(4096, kernel_size=(8,8), activation="relu"))
model2.add(Conv2D(4096, kernel_size=(1,1), activation="relu"))
model2.add(Conv2D(16, kernel_size=(1,1), activation="softmax"))
model = applications.VGG16(weights='imagenet', include_top=False, input_shape=inputshape)
F2model = Model(inputs=model.input, outputs=model2(model.output))
for layer in F2model.layers[:25]:
layer.trainable = False
F2model.compile(optimizer=optimizers.Adam(), loss="binary_crossentropy", metrics=["accuracy"])
batch_size = 128
trainsize = 36000
validsize = 12000
F2model.fit_generator(
train_generator,
steps_per_epoch=trainsize // batch_size,
epochs=5,
validation_data=valid_generator,
validation_steps=validsize // batch_size,callbacks=[tensorboard_callback])
我使用FC层训练常规网络,运行良好,但当我运行上述操作时,我得到以下错误:
ValueError Traceback (most recent call last)in <module>
4 epochs=5,
5 validation_data=valid_generator,
----> 6 validation_steps=validsize // batch_size,callbacks=[tensorboard_callback])
ValueError: Error when checking target: expected sequential_1 to have 4 dimensions, but got array with shape (32, 16)
keras的Functional API更适合解决此类问题:
model = VGG16(weights='imagenet', include_top=False, input_shape=inputshape)
x = model.output
x = Conv2D(4096, kernel_size=(8, 8), activation="relu")(x)
x = Conv2D(4096, kernel_size=(1, 1), activation="relu")(x)
out = Conv2D(16, kernel_size=(1, 1), activation="softmax")(x)
F2model = Model(inputs=model.inputs, outputs=out)
for layer in F2model.layers[:25]:
layer.trainable = False
此外,我发现您正在将二进制交叉熵与softmax激活一起使用,这可能会导致一些问题:-使用softmax和分类交叉熵
-使用sigmoid和binary_交叉熵 注意这个模型,使用4096的卷积将使你的参数数量变得非常拥挤 (本例为1.65亿) 编辑 看起来您的问题只来自标签阵列:
- 您的最后一层是卷积层,因此它期望4D阵列具有形状
,但您给它的是形状(批次大小、高度、宽度、通道)
(批次大小,16)
- 因此,将最后一层更改为:
out=density(16,activation=“softmax”)(x)
- 或者将标签数组更改为卷积层可以接受
train_generator=datagen.flow_from_dataframe(dataframe=traindf,directory="data_final",x_col="path",y_col="label",subset="training",batch_size=32,seed=42,shuffle=True,class_mode="categorical",target_size=(256,256))
valid_generator=datagen.flow_from_dataframe(dataframe=traindf,directory="data_final",x_col="path",y_col="label",subset="validation",batch_size=32,seed=42,shuffle=True,class_mode="categorical",target_size=(256,256))
if K.image_data_format()=="channels_first":
inputshape=(3,imrows,imcols)
else:
inputshape=(imrows,imcols,3)
model = VGG16(weights='imagenet', include_top=False, input_shape=inputshape)
x = model.output
x = Conv2D(4096, kernel_size=(8, 8), activation="relu")(x)
x = Conv2D(4096, kernel_size=(1, 1), activation="relu")(x)
out = Conv2D(16, kernel_size=(1, 1), activation="softmax")(x)
F2model = Model(inputs=model.inputs, outputs=out)
for layer in F2model.layers[:25]:
layer.trainable = False