跳转至

弱监督概念瓶颈学习：实现鲁棒的两阶段以物体为中心的视觉推理

文章背景与核心概要

两阶段神经符号架构通过将感知（提取预定义符号）与关系推理分离，为视觉问题求解提供了强大的框架。然而，将这些高级谓词锚定到视觉帧通常需要昂贵的标注。为了解决这一痛点，本文作者引入了动态正交概念瓶颈（Dynamic Orthogonal Concept Bottleneck, D-OCB）——这是一个以物体为中心的 slot-VAE 框架，旨在极弱监督下提取与人类对齐的符号谓词。

D-OCB 框架的核心亮点包括：在训练过程中自动学习最优的超参数分配，免去了繁琐的手动损失平衡调优；通过惩罚概念子空间之间的相关性并结合标准重建自监督，注入了概念类别独立性的先验知识；通过动态维度分配机制缓解极低监督状态下的不稳定性，允许表达充分的概念将潜在维度让渡给表现滞后的概念，从而有效防止表征坍塌并提升概念准确率；在极少标签预算下展现出强大的经验性能，实现了高概念对齐度与下游推理准确率，能够匹配甚至超越端到端范式。

摘要 (Abstract)

Two-stage neuro-symbolic architectures provide an elegant paradigm for visual problem solving by cleanly separating connectionist perception of predefined symbols from possibly later defined relational reasoning thereon. However, anchoring high-level predicates into visual frames typically necessitates annotations that are expensive to acquire. In this work, we introduce the Dynamic Orthogonal Concept Bottleneck (D-OCB), an object-centric slot-VAE framework designed to extract human-aligned symbolic predicates under extremely weak supervision. D-OCB eliminates the arduous manual tuning of loss-balancing coefficients by dynamically learning optimal hyperparameter allocations during training. To infuse prior knowledge on independence of concept categories, in addition to standard reconstruction self-supervision we penalize correlation across concept subspaces. Crucially, to combat the instability of very low supervision regimes, D-OCB incorporates a dynamic dimensionality allocation mechanism; this adaptive formulation allows well-represented concepts to yield latent dimensions to underperforming concepts that are lagging behind, effectively preventing representation collapse and significantly improving overall concept accuracy. Through an extensive empirical evaluation, we demonstrate that our framework achieves high concept alignment and downstream visual reasoning accuracy using minimal label budgets, matching or outperforming end-to-end paradigms.

两阶段神经符号架构通过清晰地将预定义符号的连接主义感知与后续可能定义的关系推理分离开来，为视觉问题求解提供了一种优雅的范式。然而，将高级谓词锚定到视觉帧中通常需要获取成本高昂的标注。在这项工作中，我们引入了动态正交概念瓶颈（D-OCB），这是一个以物体为中心的 slot-VAE 框架，旨在极弱监督下提取与人类对齐的符号谓词。D-OCB 在训练过程中动态学习最优超参数分配，消除了繁琐的手动损失平衡系数调优。为了注入关于概念类别独立性的先验知识，除了标准的重建自监督之外，我们还对概念子空间之间的相关性进行了惩罚。至关重要的是，为了克服极低监督状态下的不稳定性，D-OCB 引入了动态维度分配机制；这种自适应公式允许表征充分的概念将潜在维度让渡给表现滞后的概念，从而有效防止表征坍塌并显著提高整体概念准确率。通过广泛的实证评估，我们证明了我们的框架在极少标签预算下实现了高概念对齐度以及下游视觉推理准确率，能够匹配甚至超越端到端范式。

论文元数据 (Paper Metadata)

获取与资源 (Access & Resources)