Contents

Unsupervised


VAE

VAE TYPES

Auto Encoder

This is like a simple generative algorithm in which one model is used to reduce the data like simple compression and other is for increasing data like decompressor. This model contain different neural network model for both of task. Here there is difficult to develop or generate determined image from compressed laten space.

Variational Autoencoders

  • Limitations of autoencoders for content generation *

/vae.png


How data is generated in VAE.

https://miro.medium.com/max/2000/1*ejNnusxYrn1NRDZf4Kg2lw@2x.png
autoencoder vs VAE

https://miro.medium.com/max/1400/1*Q5dogodt3wzKKktE0v3dMQ@2x.png
LOSS FOR VAE WITH REGULARISATION

GAN

GAN TYPES