eduzhai > Applied Sciences > Engineering >

Evaluating computational models of infant phonetic learning across languages

  • king
  • (0) Download
  • 20210507
  • Save

... pages left unread,continue reading

Document pages: 7 pages

Abstract: In the first year of life, infants speech perception becomes attuned to thesounds of their native language. Many accounts of this early phonetic learningexist, but computational models predicting the attunement patterns observed ininfants from the speech input they hear have been lacking. A recent studypresented the first such model, drawing on algorithms proposed for unsupervisedlearning from naturalistic speech, and tested it on a single phone contrast.Here we study five such algorithms, selected for their potential cognitiverelevance. We simulate phonetic learning with each algorithm and perform testson three phone contrasts from different languages, comparing the results toinfants discrimination patterns. The five models display varying degrees ofagreement with empirical observations, showing that our approach can helpdecide between candidate mechanisms for early phonetic learning, and providinginsight into which aspects of the models are critical for capturing infants perceptual development.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...