PAPT++：面向单域泛化的风险感知对抗调优与生成
PAPT++: Risk-Aware Adversarial Tuning and Generation for Single Domain Generalization
浏览论文内容
中文总结 AI 辅助
PAPT++是面向单域泛化的风险感知对抗生成-训练框架，通过结合预训练文本到图像扩散模型与分布鲁棒优化，生成难分语义样本以更新分类器，在标准基准上展现优越性能。
中文摘要 AI 辅助
单域泛化（SDG）旨在从一个带标注的源域中学习模型，使其能泛化到未见过的目标域。一种常用策略是通过增广或生成样本来丰富源分布，而近期的文本到图像（T2I）扩散模型为此提供了强大的生成先验。但仅靠多样性不足以实现鲁棒泛化，因为有用的生成样本还需包含当前分类器难以处理的变异。受分布鲁棒优化（DRO）启发，我们在预训练T2I模型的类条件生成空间中定义了语义歧义集，并在其中搜索当前分类器下分类损失高的样本。为此，我们提出PAPT++，一种面向SDG的风险感知对抗生成-训练框架。PAPT++首先通过图文对齐和类内多样性正则化，为每个类别学习多样化的语义参考图像；这些参考图像随后作为分类器引导的扩散合成过程中的去噪目标，减少语义漂移的同时引导生成过程朝向具有挑战性的变异方向。生成的样本与源数据结合以更新分类器，而更新后的分类器又会反过来引导下一轮合成。通过这种方式，PAPT++逐步让分类器接触到具有挑战性但语义一致的变异。在标准SDG基准上开展的大量实验，证明了所提出的PAPT++方法的优越性及其核心组件的有效性。
英文摘要
Single domain generalization (SDG) aims to learn a model from one labeled source domain that generalizes to unseen target domains. A common strategy is to enrich the source distribution with augmented or generated samples, and recent text-to-image (T2I) diffusion models provide a strong generative prior for this purpose. However, diversity alone is insufficient for robust generalization, because useful generated samples should also capture variations that the current classifier finds difficult. Motivated by distributionally robust optimization (DRO), we define a semantic ambiguity set in the class-conditional generative space of a pretrained T2I model and search it for samples with high classification loss under the current classifier. To this end, we introduce PAPT++, a risk-aware adversarial generation-training framework for SDG. PAPT++ first learns diverse semantic reference images for each class through image-text alignment and intra-class diversity regularization. These references then serve as denoising targets during classifier-guided diffusion synthesis, reducing semantic drift while guiding generation toward challenging variations. The generated samples are combined with the source data to update the classifier, and the updated classifier guides the next synthesis round in return. In this way, PAPT++ progressively exposes the classifier to challenging yet semantically consistent variations. Extensive experiments on standard SDG benchmarks demonstrate the superiority of the proposed PAPT++ method and the effectiveness of its main components.
发表机构
- Xidian University（西安电子科技大学）
- Microsoft Research Asia（微软亚洲研究院）
机构由 AI 辅助整理，请以论文原文为准。
