Cycle-Consistent Adversarial Learning as Approximate Bayesian Inference

Louis Tiao, Edwin v. Bonilla, Fabio Ramos
July, 2018
PDF Cite Code Poster Slides Workshop Homepage
Norstedts Building on Riddarholmen in Stockholm, Sweden. © Louis Tiao

Abstract

We formalize the problem of learning interdomain correspondences in the absence of paired data as Bayesian inference in a latent variable model (LVM), where one seeks the underlying hidden representations of entities from one domain as entities from the other domain. First, we introduce implicit latent variable models, where the prior over hidden representations can be specified flexibly as an implicit distribution. Next, we develop a new variational inference (VI) algorithm for this model based on minimization of the symmetric Kullback-Leibler (KL) divergence between a variational joint and the exact joint distribution. Lastly, we demonstrate that the state-of-the-art cycle-consistent adversarial learning (CYCLEGAN) models can be derived as a special case within our proposed VI framework, thus establishing its connection to approximate Bayesian inference methods.

Type
Preprint
Publication
Presented at the ICML2018 Workshop on Theoretical Foundations and Applications of Deep Generative Models. Stockholm, Sweden, 2018.
Machine Learning Generative Adversarial Networks Unsupervised Learning Probabilistic Models Latent Variable Models Variational Inference Density Ratio Estimation
Louis Tiao
Louis Tiao
PhD Candidate (AI & Machine Learning)

Thanks for stopping by! Let’s connect – drop me a message or follow me