Variational Autoencoders(VAEs) are widely used in unsupervised learning of complicated distributions. The more classical generative models depend upon sampling techniques such as MCMC. These sampling techniques are unable to scale to high dimensional spaces, for example distribution over set of images. Due to this reason, VAEs get rid of sampling by introducing gradient based optimization.
I propose a project where I will implement a VAE framework for both feed forward and recurrent networks with rigorous testing and seamlessly integrate it with the current ann codebase of mlpack. I have kept in mind both ease of use for novice users and complete control over the VAE network for advanced users. In addition to a class, reproduction of results from the papers and a command line program will also be added. In the end, I will add a tutorial to help users get familiar with the framework.