Adaptive Methods for Real-World Domain Generalization

TL;DR

This paper introduces a domain-adaptive method leveraging test sample information for improved domain generalization, achieving state-of-the-art results and establishing a large-scale benchmark dataset.

Contribution

It proposes a novel domain-adaptive approach that uses few unlabelled test samples to construct domain embeddings, enabling better generalization to unseen domains.

Findings

Achieves state-of-the-art performance on domain generalization benchmarks.

Introduces the large-scale Geo-YFCC benchmark with 1.1 million samples.

Outperforms existing methods that do not scale or underperform on the new dataset.

Abstract

Invariant approaches have been remarkably successful in tackling the problem of domain generalization, where the objective is to perform inference on data distributions different from those used in training. In our work, we investigate whether it is possible to leverage domain information from the unseen test samples themselves. We propose a domain-adaptive approach consisting of two steps: a) we first learn a discriminative domain embedding from unsupervised training examples, and b) use this domain embedding as supplementary information to build a domain-adaptive model, that takes both the input as well as its domain into account while making predictions. For unseen domains, our method simply uses few unlabelled test examples to construct the domain embedding. This enables adaptive classification on any unseen domain. Our approach achieves state-of-the-art performance on various domain generalization benchmarks. In addition, we introduce the first real-world, large-scale domain generalization benchmark, Geo-YFCC, containing 1.1M samples over 40 training, 7 validation, and 15 test domains, orders of magnitude larger than prior work. We show that the existing approaches either do not scale to this dataset or underperform compared to the simple baseline of training a model on the union of data from all training domains. In contrast, our approach achieves a significant improvement.