Bayesian GAN

Abstract

Generative adversarial networks (GANs) can implicitly learn rich distributions over images, audio, and data which are hard to model with an explicit likelihood. We present a practical Bayesian formulation for unsupervised and semi-supervised learning with GANs. Within this framework, we use stochastic gradient Hamiltonian Monte Carlo to marginalize the weights of the generator and discriminator networks. The resulting approach is straightforward and obtains good performance without any standard interventions such as feature matching, or mini-batch discrimination. By exploring an expressive posterior over the parameters of the generator, the Bayesian GAN avoids mode-collapse, produces interpretable and diverse candidate samples, and provides state-of-the-art quantitative results for semi-supervised learning on benchmarks including SVHN, CelebA, and CIFAR-10, outperforming DCGAN, Wasserstein GANs, and DCGAN ensembles.

arXiv commentUpdated to the version that appears at Advances in Neural Information Processing Systems 30 (NIPS), 2017

Cited by in corpus (16)

• Cyclical Stochastic Gradient MCMC for Bayesian Deep Learning
• The Case for Bayesian Deep Learning
• Stabilizing Generative Adversarial Networks: A Survey
• Bayesian Variational Autoencoders for Unsupervised Out-of-Distribution Detection
• Sensor-invariant Fingerprint ROI Segmentation Using Recurrent Adversarial Learning
• Test Selection for Deep Learning Systems
• Curriculum based Dropout Discriminator for Domain Adaptation
• A Contour Stochastic Gradient Langevin Dynamics Algorithm for Simulations of Multi-modal Distributions
• Remote sensing image fusion based on Bayesian GAN
• Bayesian Cycle-Consistent Generative Adversarial Networks via Marginalizing Latent Sampling
• Generative Adversarial Networks and Adversarial Autoencoders: Tutorial and Survey
• Prb-GAN: A Probabilistic Framework for GAN Modelling
• Complexity Controlled Generative Adversarial Networks
• Stabilizing Training of Generative Adversarial Nets via Langevin Stein Variational Gradient Descent
• Exploring Dropout Discriminator for Domain Adaptation
• An Adaptive Empirical Bayesian Method for Sparse Deep Learning