Learning Disentangled Representations in the Imaging Domain

TL;DR

This paper reviews the importance, concepts, and practical methods for learning disentangled representations in imaging, highlighting their potential in computer vision and healthcare applications.

Contribution

It provides a comprehensive overview of techniques, applications, and challenges in disentangled representation learning within the imaging domain.

Findings

Disentangled representations improve generalization in imaging tasks.

Applications in medical imaging demonstrate practical benefits.

Challenges include limited supervision and domain variability.

Abstract

Disentangled representation learning has been proposed as an approach to learning general representations even in the absence of, or with limited, supervision. A good general representation can be fine-tuned for new target tasks using modest amounts of data, or used directly in unseen domains achieving remarkable performance in the corresponding task. This alleviation of the data and annotation requirements offers tantalising prospects for applications in computer vision and healthcare. In this tutorial paper, we motivate the need for disentangled representations, revisit key concepts, and describe practical building blocks and criteria for learning such representations. We survey applications in medical imaging emphasising choices made in exemplar key works, and then discuss links to computer vision applications. We conclude by presenting limitations, challenges, and opportunities.