Foresee What You Will Learn: Data Augmentation for Domain Generalization in Non-stationary Environment

TL;DR

This paper introduces a novel data augmentation method called Directional Domain Augmentation (DDA) for evolving domain generalization, enabling models to adapt to gradually shifting environments in non-stationary settings.

Contribution

The paper proposes DDA, a bi-level optimization-based data augmentation technique using meta-learning to improve domain generalization in evolving environments.

Findings

DDA outperforms existing methods on synthetic datasets.

DDA effectively models environmental evolution in real-world datasets.

The approach demonstrates superior adaptability to gradual domain shifts.

Abstract

Existing domain generalization aims to learn a generalizable model to perform well even on unseen domains. For many real-world machine learning applications, the data distribution often shifts gradually along domain indices. For example, a self-driving car with a vision system drives from dawn to dusk, with the sky darkening gradually. Therefore, the system must be able to adapt to changes in ambient illumination and continue to drive safely on the road. In this paper, we formulate such problems as Evolving Domain Generalization, where a model aims to generalize well on a target domain by discovering and leveraging the evolving pattern of the environment. We then propose Directional Domain Augmentation (DDA), which simulates the unseen target features by mapping source data as augmentations through a domain transformer. Specifically, we formulate DDA as a bi-level optimization problem and solve it through a novel meta-learning approach in the representation space. We evaluate the proposed method on both synthetic datasets and realworld datasets, and empirical results show that our approach can outperform other existing methods.