在 Keras 的顺序模型中添加 Conv2D 时出现属性错误

Question

from keras.models import Sequential
from keras.layers import Dense, Dropout, Flatten
from keras.layers import Conv2D, MaxPooling2D
from keras.optimizers import SGD

model = Sequential()
model.add(Conv2D(32, (3, 3), activation='relu', input_shape=(100, 100,3)))

输入：具有 3 个通道的 100x100 图像 32 个卷积过滤器，每个大小为 3x3 出现以下错误：

AttributeError                            Traceback (most recent call last)
<ipython-input-1-8045ebb1a70a> in <module>()
      7 # input: 100x100 images with 3 channels -> (100, 100, 3) tensors.
      8 # this applies 32 convolution filters of size 3x3 each.
----> 9 model.add(Conv2D(32, (3, 3), activation='relu', input_shape=(100, 100, 3)))

/usr/local/lib/python2.7/dist-packages/keras/models.pyc in add(self, layer)
    420                 # and create the node connecting the current layer
    421                 # to the input layer we just created.
--> 422                 layer(x)
    423 
    424             if len(layer.inbound_nodes) != 1:

/usr/local/lib/python2.7/dist-packages/keras/engine/topology.pyc in __call__(self, inputs, **kwargs)
    552 
    553             # Actually call the layer, collecting output(s), mask(s), and shape(s).
--> 554             output = self.call(inputs, **kwargs)
    555             output_mask = self.compute_mask(inputs, previous_mask)
    556 

/usr/local/lib/python2.7/dist-packages/keras/layers/convolutional.pyc in call(self, inputs)
    162                 padding=self.padding,
    163                 data_format=self.data_format,
--> 164                 dilation_rate=self.dilation_rate)
    165         if self.rank == 3:
    166             outputs = K.conv3d(

/usr/local/lib/python2.7/dist-packages/keras/backend/tensorflow_backend.pyc in conv2d(x, kernel, strides, padding, data_format, dilation_rate)
   2854     x = _preprocess_conv2d_input(x, data_format)
   2855     padding = _preprocess_padding(padding)
-> 2856     x = tf.nn.convolution(
   2857         input=x,
   2858         filter=kernel,

AttributeError: 'module' object has no attribute 'convolution'

有人可以帮忙吗？这个问题是因为新发行版keras 2.0吗？

Answer 1

将 TensorFlow 升级到 TensorFlow 1.0 是 Keras 2.0 的 先决条件。虽然在 Keras 官方网站上没有提到。

升级到 Tensorflow 1.0 即可解决上述问题

要升级 Tensorflow，请遵循此 link。

在 Keras 的顺序模型中添加 Conv2D 时出现属性错误

Getting Attribute error while adding Conv2D in sequential model of Keras

keras

keras-layer