Class-Invariant Test-Time Augmentation for Domain Generalization
This addresses domain generalization for deep learning models by providing a complementary, efficient method to improve robustness to unseen domains, though it is incremental as it builds on existing test-time strategies.
The paper tackles the problem of deep models performing poorly under distribution shifts by proposing a lightweight test-time augmentation technique called Class-Invariant Test-Time Augmentation (CI-TTA), which generates class-invariant image variants and aggregates predictions with confidence filtering, achieving consistent gains on PACS and Office-Home datasets.
Deep models often suffer significant performance degradation under distribution shifts. Domain generalization (DG) seeks to mitigate this challenge by enabling models to generalize to unseen domains. Most prior approaches rely on multi-domain training or computationally intensive test-time adaptation. In contrast, we propose a complementary strategy: lightweight test-time augmentation. Specifically, we develop a novel Class-Invariant Test-Time Augmentation (CI-TTA) technique. The idea is to generate multiple variants of each input image through elastic and grid deformations that nevertheless belong to the same class as the original input. Their predictions are aggregated through a confidence-guided filtering scheme that remove unreliable outputs, ensuring the final decision relies on consistent and trustworthy cues. Extensive Experiments on PACS and Office-Home datasets demonstrate consistent gains across different DG algorithms and backbones, highlighting the effectiveness and generality of our approach.