Question

现在我正在和CNN学习AutoEncoder。为了研究，我为MNIST数据创建了一个模型。但我无法正确设置Conv2d的输出调整。请看下面的模型图像。虽然我希望第一个Conv2d输出应为(None, 16, 28, 28)，但实际输出为(None, 1, 28, 16)。关于文档，我的代码看起来并不坏。 https://keras.io/layers/convolutional/#conv2d

你能找到我的代码的任何错误吗？

我的环境

Python 3.6.0
keras 2.0.2（后端是Tensorflow）

代码

from keras.layers import Input, Convolution2D, MaxPool2D, UpSampling2D, Conv2D
from keras.models import Model
input_img = Input(shape=(1, 28, 28))
x = Conv2D(16, (3, 3), activation='relu', padding='same')(input_img)
x = MaxPool2D((2,2), padding='same')(x)
x = Conv2D(8, (3, 3), activation='relu', padding='same')(x)
x = MaxPool2D((2,2), padding='same')(x)
x = Conv2D(8, (3, 3), activation='relu', padding='same')(x)
encoded = MaxPool2D((2,2), padding='same')(x)

x = Conv2D(8, (3,3), activation='relu', padding='same')(encoded)
x = UpSampling2D((2,2))(x)
x = Conv2D(8, (3,3), activation='relu', padding='same')(x)
x = UpSampling2D((2,2))(x)
x = Conv2D(16, (3,3), activation='relu')(x)
x = UpSampling2D((2,2))(x)
decoded = Conv2D(1, (3,3), activation='sigmoid', padding='same')(x)

autoencoder= Model(input_img, decoded)
autoencoder.compile(optimizer='adam', loss='binary_crossentropy')

from keras.utils import plot_model
plot_model(autoencoder, to_file="architecture.png", show_shapes=True)

更新

我添加了autoencoder.summary()。所以我的问题是为什么CNN的第一个输出不是(None, 16, 28, 28)？ (None, 1, 28, 16)不是我的期望。

Layer (type)                 Output Shape              Param #   
=================================================================
conv2d_181 (Conv2D)          (None, 1, 28, 16)         4048      
_________________________________________________________________
max_pooling2d_82 (MaxPooling (None, 1, 14, 16)         0         
_________________________________________________________________
conv2d_182 (Conv2D)          (None, 1, 14, 8)          1160      
_________________________________________________________________
max_pooling2d_83 (MaxPooling (None, 1, 7, 8)           0         
_________________________________________________________________
conv2d_183 (Conv2D)          (None, 1, 7, 8)           584       
_________________________________________________________________
max_pooling2d_84 (MaxPooling (None, 1, 4, 8)           0         
_________________________________________________________________
conv2d_184 (Conv2D)          (None, 1, 4, 8)           584       
_________________________________________________________________
up_sampling2d_72 (UpSampling (None, 2, 8, 8)           0         
_________________________________________________________________
conv2d_185 (Conv2D)          (None, 2, 8, 8)           584       
_________________________________________________________________
up_sampling2d_73 (UpSampling (None, 4, 16, 8)          0         
_________________________________________________________________
conv2d_186 (Conv2D)          (None, 4, 16, 16)         1168      

_________________________________________________________________
up_sampling2d_74 (UpSampling (None, 8, 32, 16)         0         
_________________________________________________________________
conv2d_187 (Conv2D)          (None, 8, 32, 1)          145       
=================================================================
Total params: 8,273.0
Trainable params: 8,273.0
Non-trainable params: 0.0
_________________________________________________________________

Updated2

我的input_img是为Theano设计的。所以我必须改变如下。否则我在image_dim_ordering

中更改了~/.keras/keras.json

# Theano style
input_img = Input(shape=(1, 28, 28))

# Tensorflow style
input_img = Input(shape=(28, 28, 1))

Answer 1

这是图像排序的常见问题。 Theano将通道维度放在形状数组的第二个元素中，如(samples, channels, width, height)，而TensorFlow将通道维度放在末尾，如(samples, width, height, channels)。您正在使用Theano订购，但后端是Tensorflow。

只需更改形状以匹配正确的顺序即可。或者，您可以在~/.keras/keras.json文件中将image_dim_ordering更改为“th”。

无法在keras中设置Conv2D的输出dim

1 个答案: