VAE with a VampPrior
This work addresses the issue of local optima and useless latent dimensions in VAEs, offering a novel prior that improves model performance for researchers in generative modeling.
The paper tackles the problem of training deep generative models by proposing a new prior called VampPrior for variational auto-encoders, which uses a mixture of variational posteriors conditioned on learnable pseudo-inputs, and shows that it achieves state-of-the-art results on six datasets in unsupervised settings.
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.