20
17

Multivariate-Information Adversarial Ensemble for Scalable Joint Distribution Matching

Abstract

A broad range of cross-mm-domain generation researches boil down to matching a joint distribution by deep generative models (DGMs). Hitherto algorithms excel in pairwise domains while as mm increases, remain struggling to scale themselves to fit a joint distribution. In this paper, we propose a domain-scalable DGM, i.e., MMI-ALI for mm-domain joint distribution matching. As an mm-domain ensemble model of ALIs \cite{dumoulin2016adversarially}, MMI-ALI is adversarially trained with maximizing Multivariate Mutual Information (MMI) w.r.t. joint variables of each pair of domains and their shared feature. The negative MMIs are upper bounded by a series of feasible losses that provably lead to matching mm-domain joint distributions. MMI-ALI linearly scales as mm increases and thus, strikes a right balance between efficacy and scalability. We evaluate MMI-ALI in diverse challenging mm-domain scenarios and verify its superiority.

View on arXiv
Comments on this paper