用球面分布原型提升图像分类可解释性
Beyond Points: Spherical Distributional Part Prototypes for Interpretable Classification
- 将每个类别建模为球面上的von Mises-Fisher混合分布，捕捉部件变异
- 在CUB-200-2011上达到最优的一致性与区分度
- 适合需要局部、非冗余解释的视觉模型可解释性研究
基于原型的神经网络通过少量部件原型实现内在可解释性。然而，现代视觉骨干网络通常在归一化的方向嵌入空间中运行，导致同一语义部件内部存在显著变异性。因此，点式原型常变得冗余或不稳定，影响解释质量和鲁棒性。本文提出vMFProto，一种分布式部件原型框架，将每类建模为超球面上的von Mises-Fisher成分混合体。每个原型学习独立的集中度，以捕捉部件特异性变异，并采用熵正则最优传输（OT）实现结构化块到原型的分配。采用两阶段训练：先通过OT驱动原型发现，再通过块级蒸馏和分布感知多样性正则化进行端到端优化。在冻结的DINO骨干网络下实验表明，vMFProto在CUB-200-2011上达到领先的稳定性和区分度，且在CUB、Stanford Dogs和Stanford Cars上获得具有竞争力的分类准确率。定性结果证实，vMFProto能生成局部化、非冗余的部件证据。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。