超越点：用于可解释分类的球形分布部分原型
Beyond Points: Spherical Distributional Part Prototypes for Interpretable Classification
AI总结：
提出vMFProto分布原型框架，用von Mises-Fisher混合建模类别，通过熵最优传输分配补丁，在CUB等数据集上实现高解释质量和竞争性准确率。
摘要：
基于原型的神经网络旨在通过依赖一小组部分原型提供内在可解释性。然而，现代视觉骨干网络通常在归一化的、方向性的嵌入空间中操作，其中每个语义部分表现出显著的类内变异性。因此，点原型常常变得冗余或不稳定，损害了解释质量和鲁棒性。我们提出vMFProto，一个分布性部分原型框架，将每个类别建模为超球面上的von Mises-Fisher混合成分。每个原型学习自己的集中度，捕获部分特定的变异性，并使用熵最优传输（OT）获得结构化的补丁到原型分配。两阶段训练方案执行OT驱动的原型发现，随后通过补丁级蒸馏和分布感知多样性正则化进行端到端精炼。在CUB-200-2011、Stanford Dogs和Stanford Cars数据集上使用冻结的DINO骨干网络的实验表明，vMFProto实现了最先进的解释质量（一致性、稳定性和独特性）以及有竞争力的准确率。定性结果证实vMFProto产生局部化、非冗余的部分证据。
英文摘要：
Prototype-based neural networks aim to provide intrinsic interpretability by grounding predictions in a small set of part prototypes. However, modern vision backbones typically operate in normalized, directional embedding spaces where each semantic part exhibits substantial intra-class variability. As a result, point prototypes often become redundant or unstable, hurting both explanation quality and robustness. We propose vMFProto, a distributional part-prototype framework that models each class as a mixture of von Mises-Fisher components on the hypersphere. Each prototype learns its own concentration, capturing part-specific variability, and we use entropic optimal transport (OT) to obtain structured patch-to-prototype assignments. A two-stage training schedule performs OT-driven prototype discovery followed by end-to-end refinement with patch-level distillation and distribution-aware diversity regularization. Experiments with frozen DINO backbones show that vMFProto achieves leading consistency and distinctiveness on CUB-200-2011 and competitive classification accuracy across CUB, Stanford Dogs, and Stanford Cars. Qualitative results confirm that vMFProto yields localized, non-redundant part evidence.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
