Feature, Alignment, and Supervision in Category Learning: A Comparative Approach with Children and Neural Networks

TL;DR

This study compares how children and convolutional neural networks learn new object categories with limited labeled data, revealing differences in generalization, biases, and the influence of feature and alignment factors.

Contribution

It provides a comparative analysis of human and neural network category learning under controlled conditions, highlighting the importance of feature, alignment, and supervision interactions.

Findings

Children generalize quickly from minimal labels but are feature-biased.

CNN performance improves with more supervision, influenced by feature and alignment.

Interactions among supervision, feature structure, and alignment are crucial for understanding learning.

Abstract

Understanding how humans and machines learn from sparse data is central to cognitive science and machine learning. Using a species-fair design, we compare children and convolutional neural networks (CNNs) in a few-shot semi-supervised category learning task. Both learners are exposed to novel object categories under identical conditions. Learners receive mixtures of labeled and unlabeled exemplars while we vary supervision (1/3/6 labels), target feature (size, shape, pattern), and perceptual alignment (high/low). We find that children generalize rapidly from minimal labels but show strong feature-specific biases and sensitivity to alignment. CNNs show a different interaction profile: added supervision improves performance, but both alignment and feature structure moderate the impact additional supervision has on learning. These results show that human-model comparisons must be drawn under the right conditions, emphasizing interactions among supervision, feature structure, and alignment rather than overall accuracy.