Feature-Critic Networks for Heterogeneous Domain Generalization

TL;DR

This paper introduces a novel learning-to-learn approach with feature-critic networks to improve domain generalization, especially in heterogeneous settings with unseen domains and categories, outperforming existing methods.

Contribution

It proposes a learned auxiliary loss via feature-critic networks for domain generalization, including challenging heterogeneous scenarios with unseen categories.

Findings

Outperforms state-of-the-art in standard domain generalization

Effective in heterogeneous domain generalization with unseen categories

Learned auxiliary loss improves generalization performance

Abstract

The well known domain shift issue causes model performance to degrade when deployed to a new target domain with different statistics to training. Domain adaptation techniques alleviate this, but need some instances from the target domain to drive adaptation. Domain generalisation is the recently topical problem of learning a model that generalises to unseen domains out of the box, and various approaches aim to train a domain-invariant feature extractor, typically by adding some manually designed losses. In this work, we propose a learning to learn approach, where the auxiliary loss that helps generalisation is itself learned. Beyond conventional domain generalisation, we consider a more challenging setting of heterogeneous domain generalisation, where the unseen domains do not share label space with the seen ones, and the goal is to train a feature representation that is useful off-the-shelf for novel data and novel categories. Experimental evaluation demonstrates that our method outperforms state-of-the-art solutions in both settings.