如何在Keras中结合LSTM和CNN模型

from keras.models import Sequential
from keras.layers import Conv2D, MaxPooling2D, LSTM, Dense, TimeDistributed, Flatten

# 假设输入数据的形状为 (num_samples, time_steps, height, width, channels)
model = Sequential()

# 添加时间分布的CNN层来处理每个时间步的图像
model.add(TimeDistributed(Conv2D(32, (3, 3), activation='relu'), input_shape=(None, 64, 64, 3)))
model.add(TimeDistributed(MaxPooling2D(pool_size=(2, 2))))
model.add(TimeDistributed(Conv2D(64, (3, 3), activation='relu')))
model.add(TimeDistributed(MaxPooling2D(pool_size=(2, 2))))
model.add(TimeDistributed(Flatten()))

# 添加LSTM层来处理时间序列信息
model.add(LSTM(128, return_sequences=False))

# 添加全连接层进行最终的分类
model.add(Dense(10, activation='softmax'))

# 编译模型
model.compile(optimizer='adam', loss='categorical_crossentropy', metrics=['accuracy'])

# 打印模型结构
model.summary()