Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Collection of a Diverse, Realistic and Annotated Dataset for Wearable Activity Recognition
School of Computing, Ulster University, Co. Antrim, Northern Ireland, United Kingdom.
School of Computing, Ulster University, Co. Antrim, Northern Ireland, United Kingdom.
School of Computing, Ulster University, Co. Antrim, Northern Ireland, United Kingdom.
Luleå University of Technology, Department of Computer Science, Electrical and Space Engineering, Computer Science.ORCID iD: 0000-0003-3191-8335
Show others and affiliations
2018 (English)In: 2018 IEEE International Conference on Pervasive Computing and Communications Workshops, PerCom Workshops 2018, IEEE, 2018, p. 555-560, article id 8480322Conference paper, Published paper (Refereed)
Abstract [en]

This paper discusses the opportunities and challenges associated with the collection of a large scale, diverse dataset for Activity Recognition. The dataset was collected by 141 undergraduate students, in a controlled environment. Students collected triaxial accelerometer data from a wearable accelerometer whilst each carrying out 3 of the 18 investigated activities, categorized into 6 scenarios of daily living. This data was subsequently labelled, anonymized and uploaded to a shared repository. This paper presents an analysis of data quality, through outlier detection and assesses the suitability of the dataset for the creation and validation of Activity Recognition models. This is achieved through the application of a range of common data driven machine learning approaches. Finally, the paper describes challenges identified during the data collection process and discusses how these could be addressed. Issues surrounding data quality, in particular, identifying and addressing poor calibration of the data were identified. Results highlight the potential of harnessing these diverse data for Activity Recognition. Based on a comparison of six classification approaches, a Random Forest provided the best classification (F-measure: 0.88). In future data collection cycles, participants will be encouraged to collect a set of 'common' activities, to support generation of a larger homogeneous dataset. Future work will seek to refine the methodology further and to evaluate model on new unseen data.

Place, publisher, year, edition, pages
IEEE, 2018. p. 555-560, article id 8480322
National Category
Media and Communication Technology
Research subject
Pervasive Mobile Computing
Identifiers
URN: urn:nbn:se:ltu:diva-72319DOI: 10.1109/PERCOMW.2018.8480322Scopus ID: 2-s2.0-85056470379ISBN: 978-1-5386-3227-7 (electronic)OAI: oai:DiVA.org:ltu-72319DiVA, id: diva2:1272240
Conference
2018 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops), 19-23 March 2018, Athens, Greece
Available from: 2018-12-18 Created: 2018-12-18 Last updated: 2018-12-18Bibliographically approved

Open Access in DiVA

No full text in DiVA

Other links

Publisher's full textScopus

Authority records BETA

Hallberg, Josef

Search in DiVA

By author/editor
Hallberg, Josef
By organisation
Computer Science
Media and Communication Technology

Search outside of DiVA

GoogleGoogle Scholar

doi
isbn
urn-nbn

Altmetric score

doi
isbn
urn-nbn
Total: 22 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf