Learning a metacognition for object perception

TL;DR

This paper introduces MetaGen, an unsupervised model that learns to represent its own perceptual reliability, enabling improved object perception by modeling how percepts are generated and identifying unreliable perceptions.

Contribution

MetaGen is the first model to jointly learn a generative metacognition of perception and object inference in an unsupervised manner, inspired by human cognitive processes.

Findings

MetaGen quickly learns effective metacognition.

MetaGen outperforms models without metacognition in accuracy.

MetaGen improves object perception in simulated environments.

Abstract

Beyond representing the external world, humans also represent their own cognitive processes. In the context of perception, this metacognition helps us identify unreliable percepts, such as when we recognize that we are seeing an illusion. Here we propose MetaGen, a model for the unsupervised learning of metacognition. In MetaGen, metacognition is expressed as a generative model of how a perceptual system produces noisy percepts. Using basic principles of how the world works (such as object permanence, part of infants' core knowledge), MetaGen jointly infers the objects in the world causing the percepts and a representation of its own perceptual system. MetaGen can then use this metacognition to infer which objects are actually present in the world. On simulated data, we find that MetaGen quickly learns a metacognition and improves overall accuracy, outperforming models that lack a metacognition.