Domain Generalization (DG) techniques have emerged as a popular approach to address the challenges of domain shift in Deep Learning (DL), with the goal of generalizing well to the target domain unseen during the training. In recent years, numerous methods have been proposed to address the DG setting, among which one popular approach is the adversarial learning-based methodology. The main idea behind adversarial DG methods is to learn domain-invariant features by minimizing a discrepancy metric. However, most adversarial DG methods use 0-1 loss based $\mathcal{H}\Delta\mathcal{H}$ divergence metric. In contrast, the margin loss-based discrepancy metric has the following advantages: more informative, tighter, practical, and efficiently optimizable. To mitigate this gap, this work proposes a novel adversarial learning DG algorithm, $\textbf{MADG}$, motivated by a margin loss-based discrepancy metric. The proposed $\textbf{MADG}$ model learns domain-invariant features across all source domains and uses adversarial training to generalize well to the unseen target domain. We also provide a theoretical analysis of the proposed $\textbf{MADG}$ model based on the unseen target error bound. Specifically, we construct the link between the source and unseen domains in the real-valued hypothesis space and derive the generalization bound using margin loss and Rademacher complexity.

Theorem 2. Consider a mixture of Theorem 3. Given the same setting as Corollary 1 and Lemma 3, for any From Lemma 3, we upper-bound the expected MDD as shown below. Using the above results and Corollary 1, we get Theorem 3 . In this section, we discuss earlier literature proposed for the DG problem, in terms of their broad categories. By conducting experiments on these benchmark datasets, we ensure a comprehensive evaluation of the proposed MADG model's performance, taking into account diverse domains, varying class distributions, and image characteristics. This section provides detailed results for each domain on all five datasets in Tables A3 to A7.

Domain Generalization (DG) techniques have emerged as a popular approach to address the challenges of domain shift in Deep Learning (DL), with the goal of generalizing well to the target domain unseen during the training. In recent years, numerous methods have been proposed to address the DG setting, among which one popular approach is the adversarial learning-based methodology. The main idea behind adversarial DG methods is to learn domain-invariant features by minimizing a discrepancy metric. However, most adversarial DG methods use 0-1 loss based \mathcal{H}\Delta\mathcal{H} divergence metric. In contrast, the margin loss-based discrepancy metric has the following advantages: more informative, tighter, practical, and efficiently optimizable.