Format

Send to

Choose Destination
PLoS One. 2018 Oct 19;13(10):e0205999. doi: 10.1371/journal.pone.0205999. eCollection 2018.

The PInSoRo dataset: Supporting the data-driven study of child-child and child-robot social dynamics.

Author information

1
Bristol Robotics Lab, University of the West of England, Bristol, United Kingdom.
2
Centre for Robotics and Neural Systems, University of Plymouth, Plymouth, United Kingdom.
3
IDLab - imec, Ghent University, Ghent, Belgium.

Abstract

The study of the fine-grained social dynamics between children is a methodological challenge, yet a good understanding of how social interaction between children unfolds is important not only to Developmental and Social Psychology, but recently has become relevant to the neighbouring field of Human-Robot Interaction (HRI). Indeed, child-robot interactions are increasingly being explored in domains which require longer-term interactions, such as healthcare and education. For a robot to behave in an appropriate manner over longer time scales, its behaviours have to be contingent and meaningful to the unfolding relationship. Recognising, interpreting and generating sustained and engaging social behaviours is as such an important-and essentially, open-research question. We believe that the recent progress of machine learning opens new opportunities in terms of both analysis and synthesis of complex social dynamics. To support these approaches, we introduce in this article a novel, open dataset of child social interactions, designed with data-driven research methodologies in mind. Our data acquisition methodology relies on an engaging, methodologically sound, but purposefully underspecified free-play interaction. By doing so, we capture a rich set of behavioural patterns occurring in natural social interactions between children. The resulting dataset, called the PInSoRo dataset, comprises 45+ hours of hand-coded recordings of social interactions between 45 child-child pairs and 30 child-robot pairs. In addition to annotations of social constructs, the dataset includes fully calibrated video recordings, 3D recordings of the faces, skeletal informations, full audio recordings, as well as game interactions.

PMID:
30339680
PMCID:
PMC6195299
DOI:
10.1371/journal.pone.0205999
[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for Public Library of Science Icon for PubMed Central
Loading ...
Support Center