Few-Shot Learning of Visual Compositional Concepts through Probabilistic Schema Induction

TL;DR

This paper introduces Probabilistic Schema Induction (PSI), a deep learning model that learns compositional visual concepts from few examples by structured analogical mapping, outperforming unstructured models and mimicking human-like learning.

Contribution

The paper presents PSI, a novel deep learning approach that performs structured analogical mapping for few-shot learning of compositional visual concepts, emphasizing the importance of relational structure.

Findings

PSI achieves human-like learning performance.

PSI outperforms unstructured prototype models.

Relational similarity increases with fewer examples.

Abstract

The ability to learn new visual concepts from limited examples is a hallmark of human cognition. While traditional category learning models represent each example as an unstructured feature vector, compositional concept learning is thought to depend on (1) structured representations of examples (e.g., directed graphs consisting of objects and their relations) and (2) the identification of shared relational structure across examples through analogical mapping. Here, we introduce Probabilistic Schema Induction (PSI), a prototype model that employs deep learning to perform analogical mapping over structured representations of only a handful of examples, forming a compositional concept called a schema. In doing so, PSI relies on a novel conception of similarity that weighs object-level similarity and relational similarity, as well as a mechanism for amplifying relations relevant to classification, analogous to selective attention parameters in traditional models. We show that PSI produces human-like learning performance and outperforms two controls: a prototype model that uses unstructured feature vectors extracted from a deep learning model, and a variant of PSI with weaker structured representations. Notably, we find that PSI's human-like performance is driven by an adaptive strategy that increases relational similarity over object-level similarity and upweights the contribution of relations that distinguish classes. These findings suggest that structured representations and analogical mapping are critical to modeling rapid human-like learning of compositional visual concepts, and demonstrate how deep learning can be leveraged to create psychological models.