VAE with a VampPrior

Jakub M. Tomczak, Max Welling
5/19/2017
cs.LGcs.AIstat.ML

Official code: jmtomczak/vae_vampprior (231 stars, MIT, last commit 2018-05-06).

Abstract

Many different methods to train deep generative models have been introduced in the past. In this paper, we propose to extend the variational auto-encoder (VAE) framework with a new type of prior which we call "Variational Mixture of Posteriors" prior, or VampPrior for short. The VampPrior consists of a mixture distribution (e.g., a mixture of Gaussians) with components given by variational posteriors conditioned on learnable pseudo-inputs. We further extend this prior to a two layer hierarchical model and show that this architecture with a coupled prior and posterior, learns significantly better models. The model also avoids the usual local optima issues related to useless latent dimensions that plague VAEs. We provide empirical studies on six datasets, namely, static and binary MNIST, OMNIGLOT, Caltech 101 Silhouettes, Frey Faces and Histopathology patches, and show that applying the hierarchical VampPrior delivers state-of-the-art results on all datasets in the unsupervised permutation invariant setting and the best results or comparable to SOTA methods for the approach with convolutional networks.

View on arXivView PDF

Code Implementations(1)

jmtomczak/vae_vamppriorOfficial100%
MIT

Code for the paper "VAE with a VampPrior", J.M. Tomczak & M. Welling

23149PythonMay 19, 20178 years agoMIT
deep-learninggenerative-modelrepresentation-learningvariational-autoencoders

Similar Papers

Instance-Adaptive Parametrization for Amortized Variational Inference

85% match2026

Towards Model-Agnostic Posterior Approximation for Fast and Accurate Variational Autoencoders

84% match2024

Discouraging posterior collapse in hierarchical Variational Autoencoders using context

84% match1 repo92023

Unbiased Learning of Deep Generative Models with Structured Discrete Representations

84% match2023

Hierarchical Residual Learning Based Vector Quantized Variational Autoencoder for Image Reconstruction and Generation

84% match2022

Lagging Inference Networks and Posterior Collapse in Variational Autoencoders

84% match1 repo1852019

Cite this paper

@article{tomczak2017with,
  title  = {VAE with a VampPrior},
  author = {Jakub M. Tomczak and Max Welling},
  year   = {2017},
  eprint = {1705.07120},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1705.07120}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1705.07120.svg)](https://codeofpaper.com/paper/1705.07120)

Discussion