Neural correlates of learned categorical perception of visual stimuli with local features

TL;DR

This study investigates how learning to categorize visual stimuli with local features induces perceptual changes and neural correlates, demonstrating that category learning can modulate perception and associated brain activity.

Contribution

It provides evidence that categorization training with local features causes perceptual and neural changes, extending previous findings with holistic features.

Findings

40% of participants learned categories successfully

Learners showed increased ERP positivity after training

Behavioral and neural changes were absent in non-learners

Abstract

Learning to categorize requires distinguishing category members from non-members by detecting the features that covary with membership. Whether this process can induce changes in perception is still a matter of debate. In prior studies, we reported Learned Categorical Perception (Learned CP) effects in the form of between-category separation and within-category compression in perceived similarity after training subjects to categorize visual stimuli with distributed, holistic features.These effects were correlated with changes in an early, perceptual component of Event Related Potentials (ERPs). Using the same methodology, in this experiment we trained 96 subjects to sort line drawings of fish with local, verbalizable features into two categories by trial and error with corrective feedback. We tested for Learned CP effects and their neural correlates by measuring subjects' pairwise dissimilarity judgments and(ERPs) before and after the training. With the same frequency of trials and feedback,about 40% of the participants succeeded in learning the categories ("learners") while the rest did not ("non-learners"). Learners showed a) significant between-category separation and within-category compression after training compared to before and b) an increase in a late parietal ERP positivity (LPC; 400-600 ms) and an early, occipital positivity (P1; 80-140 ms), correlated with categorization accuracy and degree of Learned CP. These behavioral and neural changes after learning a category, absent in Non Learners, provide further evidence that category learning can modulate perceptual processes, regardless of the nature of the visual features. We complement our experimental results with a neural net model using the same stimuli.