Learning Differentiable Logic Programs for Abstract Visual Reasoning
This addresses the challenge of enabling intelligent agents to perform reasoning using analogies on abstract concepts in visual scenarios, which is incremental as it builds on differentiable forward reasoning with memory efficiency improvements.
The paper tackled the problem of abstract visual reasoning by proposing NEUMANN, a graph-based differentiable forward reasoner that efficiently handles structured programs with functors, and it outperformed neural, symbolic, and neuro-symbolic baselines on tasks including a new behind-the-scenes scenario.
Visual reasoning is essential for building intelligent agents that understand the world and perform problem-solving beyond perception. Differentiable forward reasoning has been developed to integrate reasoning with gradient-based machine learning paradigms. However, due to the memory intensity, most existing approaches do not bring the best of the expressivity of first-order logic, excluding a crucial ability to solve abstract visual reasoning, where agents need to perform reasoning by using analogies on abstract concepts in different scenarios. To overcome this problem, we propose NEUro-symbolic Message-pAssiNg reasoNer (NEUMANN), which is a graph-based differentiable forward reasoner, passing messages in a memory-efficient manner and handling structured programs with functors. Moreover, we propose a computationally-efficient structure learning algorithm to perform explanatory program induction on complex visual scenes. To evaluate, in addition to conventional visual reasoning tasks, we propose a new task, visual reasoning behind-the-scenes, where agents need to learn abstract programs and then answer queries by imagining scenes that are not observed. We empirically demonstrate that NEUMANN solves visual reasoning tasks efficiently, outperforming neural, symbolic, and neuro-symbolic baselines.