Exciting news! Gradient has launched a FREE GPU plan. Read More
Project Details

Build an Autoencoder from Scratch in Keras

Build a VAE in Keras that can encode and decode images

The Gradient Team


VAEs are a type of convolutional neural network used for converting data from a high-dimensional space into a low-dimensional one, and then reconstructing it.

In this project we'll see how to build a variational autoencoder (VAE) from scratch in Keras. The models for the encoder, decoder, and the full VAE are saved and then loaded for testing.

For a more detailed breakdown of the code, check out How to Build a Variational Autoencoder in Keras on the blog.