Computational modeling of early language learning from acoustic speech and audiovisual input without linguistic priors

TL;DR

This paper reviews recent computational models that learn early language from speech and audiovisual input without relying on linguistic priors, highlighting advances in self-supervised and grounded learning.

Contribution

It introduces how modern models are increasingly capable of learning speech features without linguistic priors and links these models to theories of infant language development.

Findings

Models can learn speech features without linguistic priors

Shared learning principles explain early language development

Simulations are becoming more realistic and empirically grounded

Abstract

Learning to understand speech appears almost effortless for typically developing infants, yet from an information-processing perspective, acquiring a language from acoustic speech is an enormous challenge. This chapter reviews recent developments in using computational models to understand early language acquisition from speech and audiovisual input. The focus is on self-supervised and visually grounded models of perceptual learning. We show how these models are becoming increasingly powerful in learning various aspects of speech without strong linguistic priors, and how many features of early language development can be explained through a shared set of learning principles-principles broadly compatible with multiple theories of language acquisition and human cognition. We also discuss how modern learning simulations are gradually becoming more realistic, both in terms of input data and in linking model behavior to empirical findings on infant language development.