ValueError: Layer model_16期望2个输入，但它收到1个输入张量

Question

我试图使用Concatenate()来创建VGG16和VGG19的集合。我的图像的形状是(224,224,3).我不明白这个错误是关于什么。

代码如下:

# Preprocessing the Training set
train_datagen = ImageDataGenerator(rescale = 1./255,
shear_range = 0.2,
zoom_range = 0.2,
horizontal_flip = True)
# Preprocessing the Train set
training_set = train_datagen.flow_from_directory('/content/drive/MyDrive/Model Development /tbdataset/Train',
target_size = (224, 224),
batch_size = 32,
class_mode = 'categorical')
# Preprocessing the Test set
test_datagen = ImageDataGenerator(rescale = 1./255)
test_set = test_datagen.flow_from_directory('/content/drive/MyDrive/Model Development /tbdataset/Test',
target_size = (224, 224),
batch_size = 32,
class_mode = 'categorical',
shuffle=False)
vgg19 = VGG19(input_shape=IMAGE_SIZE, weights='imagenet', include_top=False)
for layer in vgg19.layers:
layer._name = layer._name + str('_19')
layer.trainable = False
vgg16 = VGG16(input_shape=IMAGE_SIZE, weights='imagenet', include_top=False)
for layer in vgg16.layers:
layer._name = layer._name + str('_16')
layer.trainable = False
vgg16_x = Flatten()(vgg16.output)
vgg19_x = Flatten()(vgg19.output)
x = Concatenate()([vgg16_x, vgg19_x])
out = Dense(2, activation='softmax')(x)
model = Model(inputs = [vgg16.input, vgg19.input], outputs = out)
model.compile(
loss='categorical_crossentropy',
optimizer=tf.keras.optimizers.Adam(
learning_rate=0.0005,
name="Adam"),
metrics=['accuracy',
'AUC',
'Precision',
'Recall',
]
)
model.summary()
r = model.fit(
training_set,
validation_data=test_set,
epochs=20,
steps_per_epoch=len(training_set),
validation_steps=len(test_set)
)

我得到以下错误:

ValueError: Layer model_16 expects 2 input(s), but it received 1 input tensors. Inputs received: [<tf.Tensor 'IteratorGetNext:0' shape=(None, None, None, None) dtype=float32>]

有谁可以指导我上面的问题吗?提前感谢!

Answer 1

如果vgg16和vgg19接收相同的输入，您可以为两者使用共享输入层。这样，你的模型将只有一个输入。

代码如下:

IMAGE_SIZE = (224,224,3)
vgg19 = tf.keras.applications.vgg19.VGG19(
input_shape=IMAGE_SIZE, weights='imagenet', include_top=False)
for layer in vgg19.layers:
layer._name = layer._name + str('_19')
layer.trainable = False
vgg16 = tf.keras.applications.vgg16.VGG16(
input_shape=IMAGE_SIZE, weights='imagenet', include_top=False)
for layer in vgg16.layers:
layer._name = layer._name + str('_16')
layer.trainable = False
inp = Input(IMAGE_SIZE)

vgg16_x = Flatten()(vgg16(inp))
vgg19_x = Flatten()(vgg19(inp))
x = Concatenate()([vgg16_x, vgg19_x])
out = Dense(2, activation='softmax')(x)
model = Model(inputs = inp, outputs = out)
model.compile(
loss='categorical_crossentropy',
optimizer=tf.keras.optimizers.Adam(
learning_rate=0.0005,
name="Adam"),
metrics=['accuracy',
'AUC',
'Precision',
'Recall',
]
)
model.summary()