DIVA: Domain Invariant Variational Autoencoders

TL;DR

DIVA is a novel generative model that learns separate latent spaces for domain, class, and residual variations to improve domain generalization, especially when unlabeled data is available, demonstrated on image datasets.

Contribution

The paper introduces DIVA, a variational autoencoder with three independent latent subspaces for domain, class, and residuals, enabling effective domain generalization with unlabeled data.

Findings

DIVA's subspaces are complementary and well-separated.

It outperforms recent methods on rotated MNIST and malaria datasets.

Incorporating unlabeled data enhances model performance.

Abstract

We consider the problem of domain generalization, namely, how to learn representations given data from a set of domains that generalize to data from a previously unseen domain. We propose the Domain Invariant Variational Autoencoder (DIVA), a generative model that tackles this problem by learning three independent latent subspaces, one for the domain, one for the class, and one for any residual variations. We highlight that due to the generative nature of our model we can also incorporate unlabeled data from known or previously unseen domains. To the best of our knowledge this has not been done before in a domain generalization setting. This property is highly desirable in fields like medical imaging where labeled data is scarce. We experimentally evaluate our model on the rotated MNIST benchmark and a malaria cell images dataset where we show that (i) the learned subspaces are indeed complementary to each other, (ii) we improve upon recent works on this task and (iii) incorporating unlabelled data can boost the performance even further.