Modeling Gestalt Visual Reasoning on the Raven's Progressive Matrices Intelligence Test Using Generative Image Inpainting Techniques

TL;DR

This paper explores how generative image inpainting models can simulate Gestalt visual reasoning in Raven's Progressive Matrices, demonstrating that perceptual processes contribute to human-like intelligence test performance.

Contribution

It introduces a novel approach using self-supervised inpainting models trained on real-world images to model perceptual reasoning in intelligence testing.

Findings

Inpainting model achieved 27/36 on Colored Progressive Matrices.

Models trained on real-world object images perform better than those trained on other datasets.

Results suggest visual regularities learned from natural images aid in reasoning about test stimuli.

Abstract

Psychologists recognize Raven's Progressive Matrices as a very effective test of general human intelligence. While many computational models have been developed by the AI community to investigate different forms of top-down, deliberative reasoning on the test, there has been less research on bottom-up perceptual processes, like Gestalt image completion, that are also critical in human test performance. In this work, we investigate how Gestalt visual reasoning on the Raven's test can be modeled using generative image inpainting techniques from computer vision. We demonstrate that a self-supervised inpainting model trained only on photorealistic images of objects achieves a score of 27/36 on the Colored Progressive Matrices, which corresponds to average performance for nine-year-old children. We also show that models trained on other datasets (faces, places, and textures) do not perform as well. Our results illustrate how learning visual regularities in real-world images can translate into successful reasoning about artificial test stimuli. On the flip side, our results also highlight the limitations of such transfer, which may explain why intelligence tests like the Raven's are often sensitive to people's individual sociocultural backgrounds.