如何使用keras/tensorflow构建无监督CNN模型？

import tensorflow as tf
from tensorflow.keras.layers import Conv2D, MaxPooling2D, UpSampling2D
from tensorflow.keras.models import Sequential

encoder = Sequential()
encoder.add(Conv2D(filters=16, kernel_size=(3, 3), activation='relu', padding='same', input_shape=(width, height, channels)))
encoder.add(MaxPooling2D(pool_size=(2, 2)))
encoder.add(Conv2D(filters=8, kernel_size=(3, 3), activation='relu', padding='same'))
encoder.add(MaxPooling2D(pool_size=(2, 2)))
encoder.add(Conv2D(filters=8, kernel_size=(3, 3), activation='relu', padding='same'))

decoder = Sequential()
decoder.add(Conv2D(filters=8, kernel_size=(3, 3), activation='relu', padding='same'))
decoder.add(UpSampling2D(size=(2, 2)))
decoder.add(Conv2D(filters=8, kernel_size=(3, 3), activation='relu', padding='same'))
decoder.add(UpSampling2D(size=(2, 2)))
decoder.add(Conv2D(filters=16, kernel_size=(3, 3), activation='relu'))
decoder.add(Conv2D(filters=channels, kernel_size=(3, 3), activation='sigmoid', padding='same'))

autoencoder = Sequential([encoder, decoder])

autoencoder.compile(optimizer='adam', loss='binary_crossentropy')
autoencoder.fit(x_train, x_train, epochs=10, batch_size=128, validation_data=(x_test, x_test))