PixelGAN Autoencoders

Abstract

In this paper, we describe the "PixelGAN autoencoder", a generative autoencoder in which the generative path is a convolutional autoregressive neural network on pixels (PixelCNN) that is conditioned on a latent code, and the recognition path uses a generative adversarial network (GAN) to impose a prior distribution on the latent code. We show that different priors result in different decompositions of information between the latent code and the autoregressive decoder. For example, by imposing a Gaussian distribution as the prior, we can achieve a global vs. local decomposition, or by imposing a categorical distribution as the prior, we can disentangle the style and content information of images in an unsupervised fashion. We further show how the PixelGAN autoencoder with a categorical prior can be directly used in semi-supervised settings and achieve competitive semi-supervised classification results on the MNIST, SVHN and NORB datasets.

Cited by in corpus (9)

• Deep Bayesian Reward Learning from Preferences
• ODIN: Automated Drift Detection and Recovery in Video Analytics
• Adversarially Approximated Autoencoder for Image Generation and Manipulation
• DeltaGAN: Towards Diverse Few-shot Image Generation with Sample-Specific Delta
• Learning Disentangled Representations with Latent Variation Predictability
• InfoVAEGAN : learning joint interpretable representations by information maximization and maximum likelihood
• Variational Mutual Information Maximization Framework for VAE Latent Codes with Continuous and Discrete Priors
• Stacked Wasserstein Autoencoder
• Learning Improved Representations by Transferring Incomplete Evidence Across Heterogeneous Tasks