Python 基于LSTM的人脸图像识别

我尝试使用基于深度学习的模型LSTM实现面部情绪识别，但没有成功。使用python 3.7 anaconda3。此代码在CNN模型上工作

from keras.preprocessing.image import ImageDataGenerator
train_datagen = ImageDataGenerator(rescale = 1./255, shear_range = 0.2, zoom_range = 0.2, horizontal_flip = True)
test_datagen = ImageDataGenerator(rescale = 1./255)

training_set = train_datagen.flow_from_directory('D:\\Datasets\\gender CK+\\training_set',
target_size = (64, 64), batch_size = 32, class_mode = 'categorical')

test_set = test_datagen.flow_from_directory('D:\\Datasets\\gender CK+\\test_set',
target_size = (64, 64), batch_size = 32, class_mode = 'categorical')

seq = Sequential()
seq.add(ConvLSTM2D(filters=32, kernel_size=(3, 3),input_shape=(None, 64, 64, 40),
    padding='same', return_sequences=True))seq.add(Batch Normalization())
seq.add(Dense(2,activation="sigmoid"))
seq.compile(loss='binary_crossentropy', optimizer='rmsprop',metrics=['accuracy'])
seq.summary()
seq.fit_generator(training_set, steps_per_epoch = 100, epochs = 4, validation_data = test_set, validation_steps = 200)

出现以下错误

ValueError：检查输入时出错：预期的conv_lst_m2d_4_输入 有5个维度，但得到了形状为（5,64,64,3）的数组

---

为什么要使用LSTM？LSTM用于序列或时间序列数据，您似乎没有。为什么要使用LSTM？LSTM用于序列或时间序列数据，您似乎没有。