mlosch/CVAE-Torch

Implementation of Convolutional Variational Auto-Encoder in Torch7

##Convolutional Variational Auto-encoder

This implementation of a CVAE is built upon the code from

• https://github.com/y0ast/VAE-Torch and
• https://github.com/soumith/dcgan.torch

and the great papers by Kingma and Welling Stochastic Gradient VB and the Variational Auto-Encoder as well as Radford, Metz and Chintala Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks.

See the github dcgan.torch project for instructions on how to run it. If you have your data ready a simple call: DATA_ROOT=<data/folder> th main_cvae.lua should suffice.

The code is MIT licensed.