Repository logo
 

Computational Audio Modelling for Robot-Assisted Assessment of Children’s Mental Wellbeing

cam.depositDate2022-10-24
cam.issuedOnline2023-02-02
cam.orpheus.counter9
cam.orpheus.successTue Feb 14 10:21:06 GMT 2023 - This item is covered by RRS with an embargo. The item is now published and embargo has been lifted.
dc.contributor.authorAbbasi, NI
dc.contributor.authorSpitale, M
dc.contributor.authorAnderson, J
dc.contributor.authorFord, T
dc.contributor.authorJones, PB
dc.contributor.authorGunes, H
dc.date.accessioned2022-10-25T23:30:32Z
dc.date.available2022-10-25T23:30:32Z
dc.date.issued2022
dc.date.updated2022-10-24T17:05:10Z
dc.description.abstractRobots endowed with the capability of assessing the mental wellbeing of children have a great potential to promote their mental health. However, very few works have explored the computational modelling of children’s mental wellbeing, which remains an open research challenge. This paper presents the first attempt to computationally assess children’s wellbeing during child-robot interactions via audio analysis. We collected a novel dataset of 26 children (8-13 y.o.) who interacted with a Nao robot to perform a verbal picture-based task. Data was collected by audio-video recording of the experiment session. The Short Mood and Feelings Questionnaire (SMFQ) was used to label the participants into two groups: (1)“higher wellbeing” (child SMFQ score <= SMFQ median), and (2) “lower wellbeing” (child SMFQ score > SMFQ median). We extracted audio features from these HRI interactions and trained and compared the performances of eight classical machine learning techniques across three cross-validation approaches: (1) 10 repetitions of 5-fold, (2) leave-one-child-out, and (3) leave-one-picture-out. We have also computed and analysed the sentiment of the audio transcriptions using the ROBERTa model. Our experimental results show that: (i) speech features are reliable for assessing children’s mental wellbeing, but they may not be sufficient on their own, and (ii) verbal information, specifically the sentiment that a picture elicited in children, may impact the children’s responses.
dc.description.sponsorshipThis work was supported by the University of Cambridge’s OHMC Small Equipment Funding. N. I. Abbasi is supported by the W.D. Armstrong Trust PhD Studentship and the Cambridge Trusts. M. Spitale and H. Gunes are supported by the EPSRC project ARoEQ under grant ref. EP/R030782/1. All research at the Department of Psychiatry in the University of Cambridge is supported by the NIHR Cambridge Biomedical Research Centre (BRC-1215-20014, particularly T. Ford) and NIHR Applied Research Collaboration East of England (P. Jones, J. Anderson).
dc.identifier.doi10.17863/CAM.89924
dc.identifier.eissn1611-3349
dc.identifier.isbn9783031246692
dc.identifier.issn0302-9743
dc.identifier.urihttps://www.repository.cam.ac.uk/handle/1810/342506
dc.language.isoeng
dc.publisherSpringer Nature Switzerland
dc.publisher.departmentDepartment of Computer Science and Technology
dc.publisher.urlhttp://dx.doi.org/10.1007/978-3-031-24670-8_3
dc.rightsAttribution 4.0 International
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/
dc.subject46 Information and Computing Sciences
dc.subject4608 Human-Centred Computing
dc.subjectMental Health
dc.subjectPediatric
dc.subjectBrain Disorders
dc.subjectBehavioral and Social Science
dc.subjectMental health
dc.subject3 Good Health and Well Being
dc.titleComputational Audio Modelling for Robot-Assisted Assessment of Children’s Mental Wellbeing
dc.typeConference Object
dcterms.dateAccepted2022-10-14
prism.publicationNameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
pubs.conference-finish-date2022-12-16
pubs.conference-name14th International Conference on Social Robotics
pubs.conference-start-date2022-12-13
pubs.funder-project-idEngineering and Physical Sciences Research Council (EP/R030782/1)
pubs.funder-project-idNational Institute for Health and Care Research (NIHR200177)
pubs.funder-project-idNational Institute for Health and Care Research (IS-BRC-1215-20014)
pubs.licence-display-nameApollo Repository Deposit Licence Agreement
pubs.licence-identifierapollo-deposit-licence-2-1
rioxxterms.versionAM
rioxxterms.versionofrecord10.1007/978-3-031-24670-8_3

Files

Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
ICSR_2022_Main_final.pdf
Size:
7.08 MB
Format:
Adobe Portable Document Format
Description:
Accepted version
Licence
https://creativecommons.org/licenses/by/4.0/