Multi-Granularity Modularized Network for Abstract Visual Reasoning
This addresses the challenge of making machines perform explainable abstract reasoning, which is incremental by building on neuro-symbolic approaches for a specific cognitive task.
The paper tackles abstract visual reasoning on the Raven Progressive Matrices Test by proposing a Multi-Granularity Modularized Network (MMoN), which learns modularized reasoning functions in a neuro-symbolic and semi-supervised way, showing it is well-suited and explainable on generalization tests.
Abstract visual reasoning connects mental abilities to the physical world, which is a crucial factor in cognitive development. Most toddlers display sensitivity to this skill, but it is not easy for machines. Aimed at it, we focus on the Raven Progressive Matrices Test, designed to measure cognitive reasoning. Recent work designed some black-boxes to solve it in an end-to-end fashion, but they are incredibly complicated and difficult to explain. Inspired by cognitive studies, we propose a Multi-Granularity Modularized Network (MMoN) to bridge the gap between the processing of raw sensory information and symbolic reasoning. Specifically, it learns modularized reasoning functions to model the semantic rule from the visual grounding in a neuro-symbolic and semi-supervision way. To comprehensively evaluate MMoN, our experiments are conducted on the dataset of both seen and unseen reasoning rules. The result shows that MMoN is well suited for abstract visual reasoning and also explainable on the generalization test.