Hands-On Unsupervised Learning with Python

Adding noise to the deep convolutional autoencoder

In this example, we are going to modify the previously developed deep convolutional autoencoder, in order to manage noisy input samples. The DAG is almost equivalent, with the difference that, now, we need to feed both noisy and original images:

import tensorflow as tfwith graph.as_default():    input_images_xl = tf.placeholder(tf.float32,                                      shape=(None, X_train.shape[1], X_train.shape[2], 1))    input_noisy_images_xl = tf.placeholder(tf.float32,                                            shape=(None, X_train.shape[1], X_train.shape[2], 1))        input_images = tf.image.resize_images(input_images_xl, (width, height),                                           method=tf.image.ResizeMethod.BICUBIC)    input_noisy_images = tf.image.resize_images(input_noisy_images_xl, (width, height),  method=tf.image.ResizeMethod.BICUBIC) ...