Uncertainty-guided Contrastive Learning for Single Source Domain Generalisation
This addresses the problem of domain shift for machine learning models trained on a single domain, enabling better generalization to unseen domains, though it is incremental as it builds on existing single-DG methods.
The paper tackles single source domain generalization by introducing CUDGNet, which uses a fictitious domain generator and contrastive learning to enhance source capacity and learn domain-invariant representations, achieving up to 7.08% improvement over state-of-the-art methods on two datasets.
In the context of single domain generalisation, the objective is for models that have been exclusively trained on data from a single domain to demonstrate strong performance when confronted with various unfamiliar domains. In this paper, we introduce a novel model referred to as Contrastive Uncertainty Domain Generalisation Network (CUDGNet). The key idea is to augment the source capacity in both input and label spaces through the fictitious domain generator and jointly learn the domain invariant representation of each class through contrastive learning. Extensive experiments on two Single Source Domain Generalisation (SSDG) datasets demonstrate the effectiveness of our approach, which surpasses the state-of-the-art single-DG methods by up to $7.08\%$. Our method also provides efficient uncertainty estimation at inference time from a single forward pass through the generator subnetwork.