发表机构
Korea Aerospace University（韩国航空航天大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
该研究针对语义分割领域基于决策的黑盒稀疏攻击研究不足的问题，提出以类别为中心的SegPAR框架，引入差异奖励提升效率，实验显示其性能优于黑盒基线且与白盒攻击相当。
AI 中文摘要
尽管基于决策的黑盒稀疏攻击具有实际应用价值，但它们在语义分割领域受到的关注有限。为了填补这一空白，我们将分类领域最具代表性的基于决策的黑盒稀疏攻击适配为基线，为这一未被充分探索的场景建立了严格的基准。在此背景下，我们证明现有方法之一因采用以图像为中心的像素累积策略而存在严重的查询效率问题，该策略会在广阔的图像空间中快速耗尽查询预算。为克服这一问题，我们提出了SegPAR，这是一种新型的基于决策的框架，采用以类别为中心的探索范式。此外，为消除像素累积过程中标准决策奖励产生的误导性反馈，我们引入了一种新型的差异奖励。大量实验表明，SegPAR在稀疏性效率和MIoU降低方面显著优于黑盒基线，同时与白盒稀疏攻击具有竞争力。代码可在提供的https URL获取。
英文摘要
Despite the practical relevance of sparse decision-based black-box threats, they have received limited attention in semantic segmentation. To bridge this gap, we adapt the most representative decision-based black-box sparse attacks from the classification domain to serve as baselines, establishing a rigorous benchmark for this underexplored setting. In this context, we demonstrate that one of the existing methods suffers from severe query inefficiency due to its image-centric pixel accumulation, which rapidly exhausts query budgets across the vast image space. To overcome this, we propose SegPAR, a novel decision-based framework that shifts to a class-centric exploration paradigm. Furthermore, to eliminate the misleading feedback generated by standard decision rewards during pixel accumulation, we introduce a novel discrepancy reward. Extensive experiments show that SegPAR significantly outperforms black-box baselines in sparsity efficiency and MIoU reduction, while remaining competitive with white-box sparse attacks. Code is available at \href{https://github.com/KAU-QuantumAILab/SegPAR}{https://github.com/KAU-QuantumAILab/SegPAR}.
CommentsECCV 2026 poster
