在 Windows 下的 R 中使用 keras 从预训练的 XCeption 架构指定卷积基时出错

Question

当我尝试使用在 ImageNet 上预训练的 XCeption 架构中的卷积基时，我在做什么真的很明显吗？这是我在问题末尾产生错误的代码：

require(keras)

conv_base1 <- application_xception(
weights = "imagenet",
include_top = FALSE,
pooling=FALSE,
input_shape = c(300, 300, 3)
)

model51 <- keras_model_sequential() %>%
conv_base1 %>%
layer_flatten() %>%
layer_dense(units = 256, activation = "relu") %>%
layer_dense(units = 1, activation = "sigmoid")

相比之下，下面使用 application_vgg16 的几乎相同的代码工作得很好：

require(keras)

conv_base2 <- application_vgg16(
weights = "imagenet",
include_top = FALSE,
pooling=FALSE,
input_shape = c(300, 300, 3)
)

model52 <- keras_model_sequential() %>%
conv_base2 %>%
layer_flatten() %>%
layer_dense(units = 2048, activation = "relu") %>%
layer_dense(units = 1, activation = "sigmoid")

我收到以下错误（在 Windows 10 x86_64-w64-mingw32/x64（64 位）上使用 R 版本 3.4.0 (2017-04-21) 使用 keras_2.1.5 R 包):

Error in py_call_impl(callable, dots$args, dots$keywords) : ValueError: Variable block1_conv1_bn_1/moving_mean/biased already exists, disallowed. Did you mean to set reuse=True in VarScope? Originally defined at:

File "D:\Anaconda3\lib\site-packages\tensorflow\python\framework\ops.py", line 1269, in init self._traceback = _extract_stack() File "D:\Anaconda3\lib\site-packages\tensorflow\python\framework\ops.py", line 2506, in create_op original_op=self._default_original_op, op_def=op_def) File "D:\Anaconda3\lib\site-packages\tensorflow\python\framework\op_def_library.py", line 767, in apply_op op_def=op_def)

Detailed traceback: File "D:\Anaconda3\lib\site-packages\keras\models.py", line 467, in add layer(x) File "D:\Anaconda3\lib\site-packages\keras\engine\topology.py", line 617, in call output = self.call(inputs, **kwargs) File "D:\Anaconda3\lib\site-packages\keras\engine\topology.py", line 2081, in call output_tensors, _, _ = self.run_internal_graph(inputs, masks) File "D:\Anaconda3\l

更多背景信息以防万一： 我在第 5.3.1 节的 "Freature extraction with data augementation" 小节中尝试用 XCeption 替换 VGG16 时运行 Chollet 和 Allaire 最优秀的书（"Fast feature extraction without data augmentation" 中的所有内容都适用于 VGG16 和 XCeption）。

Answer 1

我不明白这个错误的来源，但我怀疑它与在另一个模型中使用一个模型有关（将基本模型添加到顺序模型中）。

我建议尝试使用功能性 API 模型。但不幸的是，我不擅长 R 来理解它的符号。

思路是（copied from here，希望语法没问题，对R有更好理解的人可以修改这段代码）

首先正常定义xception模型：

conv_base1 <- application_xception(
weights = "imagenet",
include_top = FALSE,
pooling=FALSE,
input_shape = c(300, 300, 3)
)

尝试 1:

现在让我们获取该模型的输出张量并将其传递给更多层

#the inputs of the following layers is the output of the exception model     
    #this is where I can't handle with R, these two lines may be wrong
base_inputs <- conv_base1$input
base_outputs <- conv_base1$output

#the base_outputs are the input tensor to further layers
#predictions is the output tensor from those layers
predictions <- base_outputs %>%
    layer_flatten() %>% 
    layer_dense(units = 256, activation = "relu") %>% 
    layer_dense(units = 1, activation = 'sigmoid') 

# create and compile model - model starts at base_inputs and ends at predcitions
model <- keras_model(inputs = base_inputs, outputs = predictions)

尝试2：

或者，如果定义 base_inputs 和 base_outputs 不可能像在其他代码中那样定义：

inputs <- layer_input(shape = c(300,300,3))

# outputs compose input + layers, where conv_base1 should behave like a layer
predictions <- inputs %>%
    conv_base1 %>%
    layer_flatten() %>% 
    layer_dense(units = 256, activation = "relu") %>% 
    layer_dense(units = 1, activation = 'sigmoid') 

# create and compile model
model <- keras_model(inputs = inputs, outputs = predictions)

在 Windows 下的 R 中使用 keras 从预训练的 XCeption 架构指定卷积基时出错

Error specifying convolutional base from pre-trained XCeption-architecture with keras in R under Windows

r

conv-neural-network

keras

tensorflow

transfer-learning