Personalized Online Training for Physical Activity monitoring using weak labels

Federico Cruciani, I Cleland, CD Nugent, P McCullagh, Synnes Kare, Hallberg Josef

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

1 Citation (Scopus)

Abstract

The use of smartphones for activity recognition is becoming common practice. Most approaches use a single pretrained classifier to recognize activities for all users. Research studies, however, have highlighted how a personalized trained classifier could provide better accuracy. Data labeling for ground truth generation, however, is a time-consuming process. The challenge is further exacerbated when opting for a personalized approach that requires user specific datasets to be labeled, making conventional supervised approaches unfeasible. In this work, we present early results on the investigation into a weakly supervised approach for online personalized activity recognition. This paper describes: (i) a heuristic to generate weak labels used for personalized training, (ii) a comparison of accuracy obtained using a weakly supervised classifier against a conventional ground truth trained classifier. Preliminary results show an overall accuracy of 87% of a fully supervised approach against a 74% with the proposed weakly supervised approach.
LanguageEnglish
Title of host publication2018 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops)
Pages567-572
DOIs
Publication statusPublished - 8 Oct 2018
Event2018 IEEE International Conference on Pervasive Computing and Communications (PerCom Workshops) - Athens, Greece
Duration: 19 Mar 201823 Mar 2018

Conference

Conference2018 IEEE International Conference on Pervasive Computing and Communications (PerCom Workshops)
CountryGreece
CityAthens
Period19/03/1823/03/18

Fingerprint

Labels
Classifiers
Monitoring
Smartphones
Labeling

Cite this

Cruciani, F., Cleland, I., Nugent, CD., McCullagh, P., Kare, S., & Josef, H. (2018). Personalized Online Training for Physical Activity monitoring using weak labels. In 2018 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops) (pp. 567-572) https://doi.org/10.1109/PERCOMW.2018.8480292
Cruciani, Federico ; Cleland, I ; Nugent, CD ; McCullagh, P ; Kare, Synnes ; Josef, Hallberg. / Personalized Online Training for Physical Activity monitoring using weak labels. 2018 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops). 2018. pp. 567-572
@inproceedings{789309e2f87548deb1d12972cc62e04c,
title = "Personalized Online Training for Physical Activity monitoring using weak labels",
abstract = "The use of smartphones for activity recognition is becoming common practice. Most approaches use a single pretrained classifier to recognize activities for all users. Research studies, however, have highlighted how a personalized trained classifier could provide better accuracy. Data labeling for ground truth generation, however, is a time-consuming process. The challenge is further exacerbated when opting for a personalized approach that requires user specific datasets to be labeled, making conventional supervised approaches unfeasible. In this work, we present early results on the investigation into a weakly supervised approach for online personalized activity recognition. This paper describes: (i) a heuristic to generate weak labels used for personalized training, (ii) a comparison of accuracy obtained using a weakly supervised classifier against a conventional ground truth trained classifier. Preliminary results show an overall accuracy of 87{\%} of a fully supervised approach against a 74{\%} with the proposed weakly supervised approach.",
author = "Federico Cruciani and I Cleland and CD Nugent and P McCullagh and Synnes Kare and Hallberg Josef",
year = "2018",
month = "10",
day = "8",
doi = "10.1109/PERCOMW.2018.8480292",
language = "English",
isbn = "978-1-5386-3227-7",
pages = "567--572",
booktitle = "2018 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops)",

}

Cruciani, F, Cleland, I, Nugent, CD, McCullagh, P, Kare, S & Josef, H 2018, Personalized Online Training for Physical Activity monitoring using weak labels. in 2018 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops). pp. 567-572, 2018 IEEE International Conference on Pervasive Computing and Communications (PerCom Workshops), Athens, Greece, 19/03/18. https://doi.org/10.1109/PERCOMW.2018.8480292

Personalized Online Training for Physical Activity monitoring using weak labels. / Cruciani, Federico; Cleland, I; Nugent, CD; McCullagh, P; Kare, Synnes; Josef, Hallberg.

2018 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops). 2018. p. 567-572.

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

TY - GEN

T1 - Personalized Online Training for Physical Activity monitoring using weak labels

AU - Cruciani, Federico

AU - Cleland, I

AU - Nugent, CD

AU - McCullagh, P

AU - Kare, Synnes

AU - Josef, Hallberg

PY - 2018/10/8

Y1 - 2018/10/8

N2 - The use of smartphones for activity recognition is becoming common practice. Most approaches use a single pretrained classifier to recognize activities for all users. Research studies, however, have highlighted how a personalized trained classifier could provide better accuracy. Data labeling for ground truth generation, however, is a time-consuming process. The challenge is further exacerbated when opting for a personalized approach that requires user specific datasets to be labeled, making conventional supervised approaches unfeasible. In this work, we present early results on the investigation into a weakly supervised approach for online personalized activity recognition. This paper describes: (i) a heuristic to generate weak labels used for personalized training, (ii) a comparison of accuracy obtained using a weakly supervised classifier against a conventional ground truth trained classifier. Preliminary results show an overall accuracy of 87% of a fully supervised approach against a 74% with the proposed weakly supervised approach.

AB - The use of smartphones for activity recognition is becoming common practice. Most approaches use a single pretrained classifier to recognize activities for all users. Research studies, however, have highlighted how a personalized trained classifier could provide better accuracy. Data labeling for ground truth generation, however, is a time-consuming process. The challenge is further exacerbated when opting for a personalized approach that requires user specific datasets to be labeled, making conventional supervised approaches unfeasible. In this work, we present early results on the investigation into a weakly supervised approach for online personalized activity recognition. This paper describes: (i) a heuristic to generate weak labels used for personalized training, (ii) a comparison of accuracy obtained using a weakly supervised classifier against a conventional ground truth trained classifier. Preliminary results show an overall accuracy of 87% of a fully supervised approach against a 74% with the proposed weakly supervised approach.

U2 - 10.1109/PERCOMW.2018.8480292

DO - 10.1109/PERCOMW.2018.8480292

M3 - Conference contribution

SN - 978-1-5386-3227-7

SP - 567

EP - 572

BT - 2018 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops)

ER -

Cruciani F, Cleland I, Nugent CD, McCullagh P, Kare S, Josef H. Personalized Online Training for Physical Activity monitoring using weak labels. In 2018 IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom Workshops). 2018. p. 567-572 https://doi.org/10.1109/PERCOMW.2018.8480292