Abstract

This study aims to examine the role of infant-directed speech (IDS) in infant phonetic learning (the ability to acquire native speech sounds), which is crucial for further language development. This involves infants' ability to identify specific acoustic cues (phonemes) to distinguish words in their language (e.g., ship vs sheep). We will utilize the Norwegian and Danish parental speech datasets to evaluate the impact of IDS on infants' phonetic learning from a different, more holistic perspective, employing deep learning models. It will provide an opportunity to assess the use of deep learning approaches in understanding the role of IDS in infants' learning. The proposed study expands on the existing knowledge by employing deep learning models to explore the functional role of IDS in infants' learning of corner vowels (/i/, /a/, and /u/), and to determine which acoustic features of IDS are most beneficial for learning. 

About the speaker

Arun Prakash Singh, Postdoctoral Fellow, Center for Multilingualism in Society across the Lifespan, University of Oslo

Free of charge - All are welcome