• VAE → original paper
• The fundamental problem that an autoencoder is trying to solve

Autoencoders

• An autoencoder is simply a neural network that is trained to perform the task of encoding and decoding an item, such that the output from this process is as close to the original item as possible.

Autoencoder Architecture

• Autoencoder is composed of two parts:
  • Encoder → compresses high-dimensional input data into a lower-dimensional embedding vector
  • Decoder → decompresses a given embedding vector back to the original domain
• Note → It is the embedding space (also called the latent space) that is the interesting part of the autoencoder.

<aside> 💡 AUTOENCODERS AS DENOISING MODELS Autoencoders can be used to clean noisy images, since the encoder learns that it is not useful to capture the position of the random noise inside the latent space in order to reconstruct the original.

</aside>

TABLE OF CONTENTS

Build an Autoencoder

Google Colab

• Using Keras and TF
• Fashion dataset → grayscale images of clothing items, 28 x 28 pixels, comes prepackaged with TF.