正如标题所述。只有在使用以下代码时才有效：
x = Flatten()(x)在卷积层和全连接层之间。
import numpy as npimport kerasfrom keras.models import Sequential, Modelfrom keras.layers import Dense, Dropout, Flatten, Inputfrom keras.layers import Conv2D, MaxPooling2Dfrom keras.optimizers import SGD# 生成虚拟数据x_train = np.random.random((100, 100, 100, 3))y_train = keras.utils.to_categorical(np.random.randint(10, size=(100, 1)), num_classes=10)#构建模型input_layer = Input(shape=(100, 100, 3))x = Conv2D(32, (3, 3), activation='relu')(input_layer)x = Dense(256, activation='relu')(x)x = Dense(10, activation='softmax')(x)model = Model(inputs=[input_layer],outputs=[x])#编译网络sgd = SGD(lr=0.01, decay=1e-6, momentum=0.9, nesterov=True)model.compile(loss='categorical_crossentropy', optimizer=sgd)#训练网络model.fit(x_train, y_train, batch_size=32, epochs=10)否则，我会收到以下错误：
Traceback (most recent call last):File "/home/michael/practice_example.py", line 44, in <module> model.fit(x_train, y_train, batch_size=32, epochs=10)File "/usr/local/lib/python2.7/dist-packages/keras/engine/training.py", line 1435, in fit batch_size=batch_size)File "/usr/local/lib/python2.7/dist-packages/keras/engine/training.py", line 1315, in _standardize_user_data exception_prefix='target')File "/usr/local/lib/python2.7/dist-packages/keras/engine/training.py", line 127, in _standardize_input_data str(array.shape))ValueError: Error when checking target: expected dense_2 to have 4 dimensions, but got array with shape (100, 10)为什么在没有flatten()层的情况下输出会有4个维度？
回答：
根据Keras文档，
Conv2D输出形状
如果data_format=’channels_first’，则为形状为(samples, filters, new_rows, new_cols)的4D张量；如果data_format=’channels_last’，则为形状为(samples, new_rows, new_cols, filters)的4D张量。由于填充，rows和cols的值可能会发生变化。
由于您使用的是channels_last，层输出的形状将是：
# shape=(100, 100, 100, 3)x = Conv2D(32, (3, 3), activation='relu')(input_layer)# shape=(100, row, col, 32)x = Flatten()(x)# shape=(100, row*col*32) x = Dense(256, activation='relu')(x)# shape=(100, 256)x = Dense(10, activation='softmax')(x)# shape=(100, 10)错误解释（已编辑，感谢@Marcin）
使用Dense层将4D张量（shape=(100, row, col, 32)）连接到2D张量（shape=(100, 256)）仍然会形成一个4D张量（shape=(100, row, col, 256)），这不是您想要的。
# shape=(100, 100, 100, 3)x = Conv2D(32, (3, 3), activation='relu')(input_layer)# shape=(100, row, col, 32)x = Dense(256, activation='relu')(x)# shape=(100, row, col, 256)x = Dense(10, activation='softmax')(x)# shape=(100, row, col, 10)当输出4D张量与目标2D张量不匹配时，就会发生错误。
这就是为什么您需要一个Flatten层来将其从4D展平到2D。
