我正在尝试在MNIST上应用VGG16。只需测试10个数据集。但是变焦需要很长时间
将numpy导入为np 导入密码 从keras.applications.vgg16导入VGG16 从scipy导入ndimage 进口喀拉拉邦 从keras.datasets导入mnist 从keras.models导入顺序 从keras.layers导入Dense，Dropout 从keras.optimizers导入Adam 从keras.layers导入Flatten 导入matplotlib.pyplot作为plt 从keras.layers导入Dense，GlobalAveragePooling2D 从keras.models导入模型
(x_train, y_train),(x_test, y_test) = keras.datasets.mnist.load_data()
x_train = np.reshape(x_train, (-1,28,28,1)).astype('float32')/255
x_test = np.reshape(x_test, (-1,28,28,1)).astype('float32')/255
y_train = keras.utils.to_categorical(y_train)
y_test = keras.utils.to_categorical(y_test)
data_slice = 10
x_train = x_train[:data_slice,:]
y_train = y_train[:data_slice,:]
x_test = x_test[:data_slice,:]
y_test = y_test[:data_slice,:]
print('after slice')
x_train = scipy.ndimage.zoom(x_train,(8))
print('after x_train zoom')
x_test = scipy.ndimage.zoom(x_test,(8))
print('after x_test zoom')
y_train = scipy.ndimage.zoom(y_train,(8))
print('after y_train zoom')
y_test = scipy.ndimage.zoom(y_test,(8))
print('after y_test zoom')
print('after zoom')
base_model=VGG16(include_top='false', weights='imagenet')
x = base_model.output`enter code here`
print(x.shape)
x = Flatten(name='flatten')(x)
x = GlobalAveragePooling2D()(x)