CNN网络的预期输入形状和实际形状的差异

Question

嗨，我有这段代码，在检查输入时似乎给了我一个错误：预期conv2d_24_input具有形状（28，28，1）但得到了形状为（1，28，28）的数组。您能否建议我是什么导致了此错误。

from keras.datasets import mnist
from keras.models import Sequential
from keras.layers.core import Dense, Dropout, Activation, Flatten
from keras.layers.convolutional import Convolution2D, MaxPooling2D
from keras.utils import np_utils
import theano
import numpy as np
import matplotlib 
import matplotlib.pyplot as plt

#batch size to train
batch_size = 50
# number of epochs to train
nb_epoch = 20
#number of output classes
nb_classes = 10

#number of epochs to train
nb_train = 20

#input image dimension
img_rows,img_cols = 28,28

#number of convolution filters
nb_filters = 32

#size of pooling area fro max pool
nb_pool = 2
#convolution kernel size
nb_conv = 3

#data shuffeled and split between train and test data
(X_train,y_train),(X_test,y_test)= mnist.load_data()

#reshape the data

X_train = X_train.reshape(X_train.shape[0], 1, img_rows, img_cols)
X_test = X_test.reshape(X_test.shape[0], 1, img_rows, img_cols)

X_train = X_train.astype('float32')
X_test = X_test.astype('float32')

X_train /= 255
X_test /= 255

# convert class vectors to binary class matrices
Y_train = np_utils.to_categorical(y_train, nb_classes)
Y_test = np_utils.to_categorical(y_test, nb_classes)

i = 100
plt.imshow(X_train[i, 0], interpolation='nearest')
print("label : ", Y_train[i,:])

model = Sequential()

model.add(Convolution2D(nb_filters, (nb_conv, nb_conv),
                        border_mode='valid',
                        input_shape=(img_rows, img_cols,1)))
convout1 = Activation('relu')
model.add(convout1)
model.add(Convolution2D(nb_filters, (nb_conv, nb_conv)))
convout2 = Activation('relu')
model.add(convout2)
model.add(MaxPooling2D(pool_size=(nb_pool, nb_pool)))
model.add(Dropout(0.5))

model.add(Flatten())
model.add(Dense(128))
model.add(Activation('relu'))
model.add(Dropout(0.5))
model.add(Dense(nb_classes))
model.add(Activation('softmax'))
model.compile(loss='categorical_crossentropy', optimizer='adadelta', metrics=['accuracy'])

#%%

hist = model.fit(X_train, Y_train, batch_size=batch_size, nb_epoch=nb_epoch,
              verbose=1, validation_data=(X_test, Y_test))


hist = model.fit(X_train, Y_train, batch_size=batch_size, nb_epoch=nb_epoch,
              verbose=1, validation_split=0.2)


#%%       

score = model.evaluate(X_test, Y_test, verbose=0)
print('Test score:', score[0])
print('Test accuracy:', score[1])
print(model.predict_classes(X_test[1:5]))
print(Y_test[1:5])

def plot_filters(layer, x,y):
    filters = layer.W.get_value()
    fig = plt.figure()
    for j in range(len(filters)):
        ax = fig.add_subplot(y,x,j+1)
        ax.matshow(filters[j][0],cmap =matplotlib.cm.binary)

        plt.xticks(np.array([]))
        plt.yticks(np.array([]))
    plt.tight_layout
    return plt

#plot the first convolution layer filters
plot_filters(model.layers[0],8,4)

#Visualising intermediate layers
output_layer = model.layers[1].get_output()
output_fn = theano.function([model.layers[0].get_input()], output_layer)

#the input imageinput_image=X_train[0:1,:,:,:]
input_image=X_train[0:1,:,:,:]
print(input_image.shape)

# Rearrange dimension so we can plot the result 
output_image = np.rollaxis(np.rollaxis(output_image, 3, 1), 3, 1)
print(output_image.shape)

fig =  plt.figure(figsize=(16,8))
for i in range(32):
    plt.subplot(1,10,i+1)
    plt.imshow(im[0,:,:,i],interpolation ='nearest') #to see the first filter
    plt.axis('off')

我怀疑此错误是在重新排列尺寸时引起的。我该如何解决。