Mikhail Surtsukov

Mikhail Surtsukov

I am Deep Learning enthusiast interested in Computer Vision, Bayesian Deep Learning and Generative Models

AutoEncoders in Keras: VAE

less than 1 minute read

In the last part, we have already discussed what hidden variables are, looked at their distribution, and also understood that it is difficult to generate new objects from the distribution of latent variables in ordinary autoencoders. In order to be able to generate new objects, the space of latent variables must be predictable.

Variational Autoencoders are autoencoders that learn to map objects into a given hidden space and sample from it. Therefore, variational autoencoders are also referred to the family of generative models.

pde Illustration from here

Full Russian text is available here

Repository with jupyter notebook

Updated:

You may also enjoy

Neural Ordinary Differential Equations

19 minute read

A significant portion of processes can be described by differential equations: let it be evolution of physical systems, medical conditions of a patient, fund...

Generative models collection

2 minute read

PyTorch implementations of various generative models to be trained and evaluated on CelebA dataset. The models are: Deep Convolutional GAN, Least Squares GAN...

AutoEncoders in Keras: VAE-GAN

less than 1 minute read

In the previous part, we created a CVAE autoencoder, whose decoder is able to generate a digit of a given label, we also tried to create pictures of numbers ...