Learning Discrete and Continuous Factors of Data via Alternating Disentanglement

TL;DR

This paper introduces a novel unsupervised method for disentangling discrete and continuous factors in data, using an alternating minimization approach that improves over existing techniques in disentanglement quality.

Contribution

The authors propose a simple, discriminator-free approach for minimizing total correlation and a separate inference procedure for discrete factors, enabling effective alternating optimization.

Findings

Successfully disentangles discrete and continuous factors

Outperforms existing methods on disentanglement metrics

Uses an alternating minimization framework for improved results

Abstract

We address the problem of unsupervised disentanglement of discrete and continuous explanatory factors of data. We first show a simple procedure for minimizing the total correlation of the continuous latent variables without having to use a discriminator network or perform importance sampling, via cascading the information flow in the $\beta$-vae framework. Furthermore, we propose a method which avoids offloading the entire burden of jointly modeling the continuous and discrete factors to the variational encoder by employing a separate discrete inference procedure. This leads to an interesting alternating minimization problem which switches between finding the most likely discrete configuration given the continuous factors and updating the variational encoder based on the computed discrete factors. Experiments show that the proposed method clearly disentangles discrete factors and significantly outperforms current disentanglement methods based on the disentanglement score and inference network classification score. The source code is available at https://github.com/snu-mllab/DisentanglementICML19.