Structured Disentangled Representations

Babak Esmaeili, Hao Wu, Sarthak Jain, Alican Bozkurt, N Siddharth, Brooks Paige, Dana H. Brooks, Jennifer Dy, Jan-Willem van de Meent

Research output: Chapter in Book/Report/Conference proceedingConference contribution


Deep latent-variable models learn representations of high-dimensional data in an unsupervised manner. A number of recent efforts have focused on learning representations that disentangle statistically independent axes of variation by introducing modifications to the standard objective function. These approaches generally assume a simple diagonal Gaussian prior and as a result are not able to reliably disentangle discrete factors of variation. We propose a two-level hierarchical objective to control relative degree of statistical independence between blocks of variables and individual variables within blocks. We derive this objective as a generalization of the evidence lower bound, which allows us to explicitly represent the trade-offs between mutual information between data and representation, KL divergence between representation and prior, and coverage of the support of the empirical data distribution. Experiments on a variety of datasets demonstrate that our objective can not only disentangle discrete variables, but that doing so also improves disentanglement of other variables and, importantly, generalization even to unseen combinations of factors.
Original languageEnglish
Title of host publicationProceedings of the 22nd International Conference on Artificial Intelligence and Statistics
EditorsKamalika Chaudhuri, Masashi Sugiyama
Number of pages10
Publication statusPublished - 18 Apr 2019
Event22nd International Conference on Artificial Intelligence and Statistics - Naha, Japan
Duration: 16 Apr 201918 Apr 2019

Publication series

NameProceedings of Machine Learning Research
ISSN (Electronic)2640-3498


Conference22nd International Conference on Artificial Intelligence and Statistics
Abbreviated titleAISTATS 2019
Internet address


Dive into the research topics of 'Structured Disentangled Representations'. Together they form a unique fingerprint.

Cite this