Domain Generalization (DG) is a critical area that focuses on developing models capable of performing well on data from unseen distributions, which is essential for real-world applications. Existing approaches primarily concentrate on learning domain-invariant features, which assume that a model robust to variations in the source domains will generalize well to unseen target domains. However, these approaches neglect a deeper analysis at the parameter level, which makes the model hard to explicitly differentiate between parameters sensitive to domain shifts and those robust, potentially hindering its overall ability to generalize. In order to address these limitations, we first build a covariance-based parameter sensitivity analysis framework to quantify the sensitivity of each parameter in a model to domain shifts. By computing the covariance of parameter gradients across multiple source domains, we can identify parameters that are more susceptible to domain variations, which serves as our theoretical foundation. Based on this, we propose Domain-Sensitive Parameter Regularization (DSP-Reg), a principled framework that guides model optimization by a soft regularization technique that encourages the model to rely more on domain-invariant parameters while suppressing those that are domain-specific. This approach provides a more granular control over the model's learning process, leading to improved robustness and generalization to unseen domains. Extensive experiments on benchmarks, such as PACS, VLCS, OfficeHome, and DomainNet, demonstrate that DSP-Reg outperforms state-of-the-art approaches, achieving an average accuracy of 66.7\% and surpassing all baselines.

翻译：领域泛化（Domain Generalization, DG）是一个关键研究方向，致力于开发能够在未见分布数据上表现良好的模型，这对于实际应用至关重要。现有方法主要集中于学习领域不变特征，其假设对源领域变化鲁棒的模型能够良好地泛化至未见目标领域。然而，这些方法缺乏在参数层面进行更深入的分析，使得模型难以明确区分对领域偏移敏感的参数与鲁棒参数，这可能阻碍其整体泛化能力。为应对这些局限性，我们首先构建了一个基于协方差的参数敏感性分析框架，以量化模型中每个参数对领域偏移的敏感性。通过计算多个源领域中参数梯度的协方差，我们可以识别更容易受领域变化影响的参数，这构成了我们的理论基础。基于此，我们提出了域敏感参数正则化（Domain-Sensitive Parameter Regularization, DSP-Reg），这是一个原则性框架，通过软正则化技术指导模型优化，鼓励模型更多地依赖领域不变参数，同时抑制领域特定参数。该方法为模型的学习过程提供了更精细的控制，从而提升了对未见领域的鲁棒性和泛化能力。在PACS、VLCS、OfficeHome和DomainNet等基准数据集上的大量实验表明，DSP-Reg优于现有先进方法，平均准确率达到66.7%，超越了所有基线模型。