Child-activity recognition from multi-sensor data

Sabri Boughorbel, Jeroen Breebaart, Fons Bruekers, Ingrid Flinsenberg, Warner Ten Kate

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

The automatic recognition of child activity using multi-sensor data enablesvarious applications such as child-development monitoring, energy-expenditure estimation, child-obesity prevention, child safety in and around the home, etc. We formulate the activity recognition task as a classification problembased on multiple sensors embedded in a wearable device. The approach we propose in this paper isto apply spectral analysistechniques of multiple sensor data for activity recognition. Quadratic Discriminant Analysis (QDA) classifieris then trained using manually annotated data and applied for activity recognition. The obtained experimental results for the recognition of 7 activities based on a limited data set are promising and show the potential of the proposed method.

Original languageEnglish
Title of host publicationSelected Papers from the Proceedings of the 7th International Conference on Methods and Techniques in Behavioral Research - Digital Edition, MB'10
DOIs
Publication statusPublished - 2011
Externally publishedYes
Event7th International Conference on Methods and Techniques in Behavioral Research, MB'10 - Eindhoven, Netherlands
Duration: 24 Aug 201027 Aug 2010

Other

Other7th International Conference on Methods and Techniques in Behavioral Research, MB'10
CountryNetherlands
CityEindhoven
Period24/8/1027/8/10

Fingerprint

Sensors
Discriminant analysis
Monitoring
Energy Metabolism

Keywords

  • Activity classification
  • Activity recognition
  • Feature extraction

ASJC Scopus subject areas

  • Human-Computer Interaction
  • Computer Networks and Communications
  • Computer Vision and Pattern Recognition
  • Software

Cite this

Boughorbel, S., Breebaart, J., Bruekers, F., Flinsenberg, I., & Ten Kate, W. (2011). Child-activity recognition from multi-sensor data. In Selected Papers from the Proceedings of the 7th International Conference on Methods and Techniques in Behavioral Research - Digital Edition, MB'10 [38] https://doi.org/10.1145/1931344.1931382

Child-activity recognition from multi-sensor data. / Boughorbel, Sabri; Breebaart, Jeroen; Bruekers, Fons; Flinsenberg, Ingrid; Ten Kate, Warner.

Selected Papers from the Proceedings of the 7th International Conference on Methods and Techniques in Behavioral Research - Digital Edition, MB'10. 2011. 38.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Boughorbel, S, Breebaart, J, Bruekers, F, Flinsenberg, I & Ten Kate, W 2011, Child-activity recognition from multi-sensor data. in Selected Papers from the Proceedings of the 7th International Conference on Methods and Techniques in Behavioral Research - Digital Edition, MB'10., 38, 7th International Conference on Methods and Techniques in Behavioral Research, MB'10, Eindhoven, Netherlands, 24/8/10. https://doi.org/10.1145/1931344.1931382
Boughorbel S, Breebaart J, Bruekers F, Flinsenberg I, Ten Kate W. Child-activity recognition from multi-sensor data. In Selected Papers from the Proceedings of the 7th International Conference on Methods and Techniques in Behavioral Research - Digital Edition, MB'10. 2011. 38 https://doi.org/10.1145/1931344.1931382
Boughorbel, Sabri ; Breebaart, Jeroen ; Bruekers, Fons ; Flinsenberg, Ingrid ; Ten Kate, Warner. / Child-activity recognition from multi-sensor data. Selected Papers from the Proceedings of the 7th International Conference on Methods and Techniques in Behavioral Research - Digital Edition, MB'10. 2011.
@inproceedings{1db5d2a684274c0cafd16a0e4025e1f1,
title = "Child-activity recognition from multi-sensor data",
abstract = "The automatic recognition of child activity using multi-sensor data enablesvarious applications such as child-development monitoring, energy-expenditure estimation, child-obesity prevention, child safety in and around the home, etc. We formulate the activity recognition task as a classification problembased on multiple sensors embedded in a wearable device. The approach we propose in this paper isto apply spectral analysistechniques of multiple sensor data for activity recognition. Quadratic Discriminant Analysis (QDA) classifieris then trained using manually annotated data and applied for activity recognition. The obtained experimental results for the recognition of 7 activities based on a limited data set are promising and show the potential of the proposed method.",
keywords = "Activity classification, Activity recognition, Feature extraction",
author = "Sabri Boughorbel and Jeroen Breebaart and Fons Bruekers and Ingrid Flinsenberg and {Ten Kate}, Warner",
year = "2011",
doi = "10.1145/1931344.1931382",
language = "English",
isbn = "9781605589268",
booktitle = "Selected Papers from the Proceedings of the 7th International Conference on Methods and Techniques in Behavioral Research - Digital Edition, MB'10",

}

TY - GEN

T1 - Child-activity recognition from multi-sensor data

AU - Boughorbel, Sabri

AU - Breebaart, Jeroen

AU - Bruekers, Fons

AU - Flinsenberg, Ingrid

AU - Ten Kate, Warner

PY - 2011

Y1 - 2011

N2 - The automatic recognition of child activity using multi-sensor data enablesvarious applications such as child-development monitoring, energy-expenditure estimation, child-obesity prevention, child safety in and around the home, etc. We formulate the activity recognition task as a classification problembased on multiple sensors embedded in a wearable device. The approach we propose in this paper isto apply spectral analysistechniques of multiple sensor data for activity recognition. Quadratic Discriminant Analysis (QDA) classifieris then trained using manually annotated data and applied for activity recognition. The obtained experimental results for the recognition of 7 activities based on a limited data set are promising and show the potential of the proposed method.

AB - The automatic recognition of child activity using multi-sensor data enablesvarious applications such as child-development monitoring, energy-expenditure estimation, child-obesity prevention, child safety in and around the home, etc. We formulate the activity recognition task as a classification problembased on multiple sensors embedded in a wearable device. The approach we propose in this paper isto apply spectral analysistechniques of multiple sensor data for activity recognition. Quadratic Discriminant Analysis (QDA) classifieris then trained using manually annotated data and applied for activity recognition. The obtained experimental results for the recognition of 7 activities based on a limited data set are promising and show the potential of the proposed method.

KW - Activity classification

KW - Activity recognition

KW - Feature extraction

UR - http://www.scopus.com/inward/record.url?scp=79952520898&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79952520898&partnerID=8YFLogxK

U2 - 10.1145/1931344.1931382

DO - 10.1145/1931344.1931382

M3 - Conference contribution

AN - SCOPUS:79952520898

SN - 9781605589268

BT - Selected Papers from the Proceedings of the 7th International Conference on Methods and Techniques in Behavioral Research - Digital Edition, MB'10

ER -