HYPO: Hyperspherical Out-of-Distribution Generalization
It addresses the critical problem of real-world deployment for machine learning models by improving OOD generalization, though it appears incremental as it builds on existing prototypical learning methods.
The paper tackles out-of-distribution generalization by proposing HYPO, a framework that learns domain-invariant representations in a hyperspherical space, and demonstrates superior performance on challenging benchmarks.
Out-of-distribution (OOD) generalization is critical for machine learning models deployed in the real world. However, achieving this can be fundamentally challenging, as it requires the ability to learn invariant features across different domains or environments. In this paper, we propose a novel framework HYPO (HYPerspherical OOD generalization) that provably learns domain-invariant representations in a hyperspherical space. In particular, our hyperspherical learning algorithm is guided by intra-class variation and inter-class separation principles -- ensuring that features from the same class (across different training domains) are closely aligned with their class prototypes, while different class prototypes are maximally separated. We further provide theoretical justifications on how our prototypical learning objective improves the OOD generalization bound. Through extensive experiments on challenging OOD benchmarks, we demonstrate that our approach outperforms competitive baselines and achieves superior performance. Code is available at https://github.com/deeplearning-wisc/hypo.