GALA：基于图扩散与拼图对齐的无源域适应方法
GALA: Graph Diffusion-based Alignment with Jigsaw for Source-free Domain Adaptation
- University of California, Los Angeles（加利福尼亚大学洛杉矶分校）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
本文提出GALA方法，通过基于分数的图扩散模型将目标图重建为源风格图，并结合课程学习伪标签与图拼图一致性学习，解决无源图域适应中的域偏移和标签稀缺问题。
AI中文摘要：
无源域适应是一个至关重要的机器学习课题，因为它在现实世界中具有大量应用，尤其是在数据隐私方面。现有方法主要关注欧几里得数据，如图像和视频，而对非欧几里得图数据的探索仍然很少。近期的图神经网络（GNN）方法在无源适应场景中可能因域偏移和标签稀缺而遭受严重的性能下降。在本研究中，我们提出了一种名为基于图扩散与拼图的对齐方法（GALA）的新方法，专门用于无源图域适应。为了实现域对齐，GALA采用图扩散模型从目标数据重建源风格图。具体来说，使用源图训练一个基于分数的图扩散模型，以学习生成式源风格。然后，我们通过随机微分方程对目标图引入扰动，而不是从先验分布中采样，随后通过逆向过程重建源风格图。我们将源风格图输入现成的GNN，并引入带有课程学习的类别特定阈值，从而为目标图生成准确且无偏的伪标签。此外，我们开发了一种简单而有效的图混合策略，称为图拼图，用于组合高置信度图和低置信度图，这可以通过一致性学习增强泛化能力和鲁棒性。在基准数据集上的大量实验验证了GALA的有效性。
英文摘要：
Source-free domain adaptation is a crucial machine learning topic, as it contains numerous applications in the real world, particularly with respect to data privacy. Existing approaches predominantly focus on Euclidean data, such as images and videos, while the exploration of non-Euclidean graph data remains scarce. Recent graph neural network (GNN) approaches can suffer from serious performance decline due to domain shift and label scarcity in source-free adaptation scenarios. In this study, we propose a novel method named Graph Diffusion-based Alignment with Jigsaw (GALA), tailored for source-free graph domain adaptation. To achieve domain alignment, GALA employs a graph diffusion model to reconstruct source-style graphs from target data. Specifically, a score-based graph diffusion model is trained using source graphs to learn the generative source styles. Then, we introduce perturbations to target graphs via a stochastic differential equation instead of sampling from a prior, followed by the reverse process to reconstruct source-style graphs. We feed the source-style graphs into an off-the-shelf GNN and introduce class-specific thresholds with curriculum learning, which can generate accurate and unbiased pseudo-labels for target graphs. Moreover, we develop a simple yet effective graph-mixing strategy named graph jigsaw to combine confident graphs and unconfident graphs, which can enhance generalization capabilities and robustness via consistency learning. Extensive experiments on benchmark datasets validate the effectiveness of GALA.
