Prosody-Based Vocal Emotional Alignment in Infant-Caregiver Interaction

Ming Li,Jiarui Li,Dan Zhang,Yukie Nagai
DOI: https://doi.org/10.1109/icdl55364.2023.10364533
2023-01-01
Abstract:Infants actively interact with caregivers to develop social competence by aligning behaviors and mental states. Previous research has demonstrated the early ability to align core affect through distinct naïve nonverbal behaviors. However, the extent to which infants align vocal emotions with caregivers more finely has remained largely unexplored. This study investigated the nuances of vocal emotions during 60 minutes of natural conversations between caregivers and their 9-month-old infants from 25 families. To this end, a pre-trained deep neural network (DNN) based on contrastive learning was applied to extract emotion representations using the prosodic features of infants' and caregivers' vocalizations. The results revealed that infants and caregivers align their vocal emotions in natural interactions. While infants showed less variability in vocal emotions than their caregivers, the variabilities of both parties were positively correlated at the family level, indicating the role of parental vocal emotions in facilitating infant vocal emotions. These findings provide insight into the early interactions of vocal emotions in family conversations and highlight the significance of the quality of caregivers' vocal expressions, i.e., the displayed emotional variability.
What problem does this paper attempt to address?