当我尝试执行以下代码时，出现以下问题。怎么解决呢？基本上，RA_unit_3是一个预处理函数，具有与输入相同的形状。但是错误普遍存在。
UNet_02V2_CR_RAU.py:403: UserWarning: Update your `Model` call to the Keras 2 API: `Model(inputs=Tensor("in..., outputs=Tensor("co...)`
model = Model(input = inputs, output = conv10)
Traceback (most recent call last):
File "UNet_02V2_CR_RAU.py", line 433, in <module>
model=unet(input_size = (1,image_height,image_width,3))
File "UNet_02V2_CR_RAU.py", line 403, in unet
model = Model(input = inputs, output = conv10)
File "/home/harin/anaconda3/envs/tfgpuenv/lib/python3.7/site-packages/keras/legacy/interfaces.py", line 91, in wrapper
return func(*args, **kwargs)
File "/home/harin/anaconda3/envs/tfgpuenv/lib/python3.7/site-packages/keras/engine/network.py", line 93, in __init__
self._init_graph_network(*args, **kwargs)
File "/home/harin/anaconda3/envs/tfgpuenv/lib/python3.7/site-packages/keras/engine/network.py", line 231, in _init_graph_network
self.inputs, self.outputs)
File "/home/harin/anaconda3/envs/tfgpuenv/lib/python3.7/site-packages/keras/engine/network.py", line 1366, in _map_graph_network
tensor_index=tensor_index)
File "/home/harin/anaconda3/envs/tfgpuenv/lib/python3.7/site-packages/keras/engine/network.py", line 1353, in build_map
node_index, tensor_index)
File "/home/harin/anaconda3/envs/tfgpuenv/lib/python3.7/site-packages/keras/engine/network.py", line 1353, in build_map
node_index, tensor_index)
File "/home/harin/anaconda3/envs/tfgpuenv/lib/python3.7/site-packages/keras/engine/network.py", line 1353, in build_map
node_index, tensor_index)
[Previous line repeated 10 more times]
File "/home/harin/anaconda3/envs/tfgpuenv/lib/python3.7/site-packages/keras/engine/network.py", line 1325, in build_map
node = layer._inbound_nodes[node_index]
AttributeError: 'NoneType' object has no attribute '_inbound_nodes'
我的主要方法如下：
def unet(pretrained_weights = None,input_size = None):
inputs = Input(batch_shape=input_size)
conv1 = Conv2D(64, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(inputs)
conv1 = Conv2D(64, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv1)
pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)
pool1 = RA_unit_3(x=pool1,h=pool1.shape[1].value, w=pool1.shape[2].value,n=16)
conv2 = Conv2D(128, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(pool1)
conv2 = Conv2D(128, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv2)
pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)
# pool2 = RA_unit_3(x=pool2,h=pool2.shape[1].value, w=pool2.shape[2].value,n=16)
conv3 = Conv2D(256, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(pool2)
conv3 = Conv2D(256, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv3)
pool3 = MaxPooling2D(pool_size=(2, 2))(conv3)
# pool3 = RA_unit_3(x=pool3,h=pool3.shape[1].value, w=pool3.shape[2].value,n=16)
conv4 = Conv2D(512, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(pool3)
conv4 = Conv2D(512, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv4)
drop4 = Dropout(0.5)(conv4)
pool4 = MaxPooling2D(pool_size=(2, 2))(drop4)
# pool4 = RA_unit_3(x=pool4,h=pool4.shape[1].value, w=pool4.shape[2].value,n=16)
conv5 = Conv2D(1024, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(pool4)
conv5 = Conv2D(1024, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv5)
drop5 = Dropout(0.5)(conv5)
up6 = Conv2D(512, 2, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(UpSampling2D(size = (2,2))(drop5))
merge6 = concatenate([drop4,up6], axis = 3)
# merge6 = RA_unit(x=merge6,n=16)
conv6 = Conv2D(512, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(merge6)
conv6 = Conv2D(512, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv6)
up7 = Conv2D(256, 2, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(UpSampling2D(size = (2,2))(conv6))
merge7 = concatenate([conv3,up7], axis = 3)
# merge7 = RA_unit(x=merge7,n=16)
conv7 = Conv2D(256, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(merge7)
conv7 = Conv2D(256, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv7)
up8 = Conv2D(128, 2, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(UpSampling2D(size = (2,2))(conv7))
merge8 = concatenate([conv2,up8], axis = 3)
# merge8 = RA_unit(x=merge8,n=16)
conv8 = Conv2D(128, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(merge8)
conv8 = Conv2D(128, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv8)
up9 = Conv2D(64, 2, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(UpSampling2D(size = (2,2))(conv8))
merge9 = concatenate([conv1,up9], axis = 3)
# merge9 = RA_unit(x=merge9,n=16)
conv9 = Conv2D(64, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(merge9)
conv9 = Conv2D(64, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv9)
conv9 = Conv2D(2, 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv9)
conv10 = Conv2D(3, 1, activation = 'sigmoid')(conv9)
model = Model(input = inputs, output = conv10)
model.compile(optimizer = Adam(lr = 1e-4), loss = 'binary_crossentropy', metrics = ['accuracy']) # original 1e-4 | 2e-4 = 0.00020
model.summary()
if(pretrained_weights):
model.load_weights(pretrained_weights)
return model
RA_unit_3方法：
def RA_unit_3(x, h, w, n):
x_1 = tf.nn.avg_pool2d(x, ksize=[1, h/n, 2, 1], strides=[1, h/n, 2, 1], padding="SAME")
x_t = tf.zeros([1, h, w, 0], tf.float32)
for k in range(n):
x_t_1 = tf.slice(x_1, [0,k,0,0], [1,1,int(w/2),x.shape[3].value])
x_t_2 = tf.image.resize(x_t_1, [h,w], 1)
x_t_3 = tf.abs(x - x_t_2)
x_t = tf.concat([x_t, x_t_3], axis=3)
x_out = tf.concat([x, x_t], axis=3)
conv = Conv2D(x.shape[3], 3, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(x_out)
conv = MaxPooling2D(pool_size=(1, 1))(conv)
return conv
下一行的输入和输出是相同的。
pool1 = RA_unit_3(x=pool1,h=pool1.shape[1].value, w=pool1.shape[2].value,n=16)