Ghent University Academic Bibliography

Advanced

Discovering activity patterns in office environment using a network of low-resolution visual sensors

Mohamed Eldib UGent, Francis Deboeverie, Wilfried Philips UGent and Hamid Aghajan UGent (2018) JOURNAL OF AMBIENT INTELLIGENCE AND HUMANIZED COMPUTING . 9(2). p.381-411
abstract
Understanding activity patterns in office environments is important in order to increase workers’ comfort and productivity. This paper proposes an automated system for discovering activity patterns of multiple persons in a work environment using a network of cheap low-resolution visual sensors (900 pixels). Firstly, the users’ locations are obtained from a robust people tracker based on recursive maximum likelihood principles. Secondly, based on the users’ mobility tracks, the high density positions are found using a bivariate kernel density estimation. Then, the hotspots are detected using a confidence region estimation. Thirdly, we analyze the individual’s tracks to find the starting and ending hotspots. The starting and ending hotspots form an observation sequence, where the user’s presence and absence are detected using three powerful Probabilistic Graphical Models (PGMs). We describe two approaches to identify the user’s status: a single model approach and a two-model mining approach. We evaluate both approaches on video sequences captured in a real work environment, where the persons’ daily routines are recorded over 5 months. We show how the second approach achieves a better performance than the first approach. Routines dominating the entire group’s activities are identified with a methodology based on the Latent Dirichlet Allocation topic model. We also detect routines which are characteristic of persons. More specifically, we perform various analysis to determine regions with high variations, which may correspond to specific events.
Please use this url to cite or link to this publication:
author
organization
year
type
journalArticle (original)
publication status
published
subject
keyword
Visual sensor network, Supervised learning, Probabilistic graphical models, Topic models, Sequence mining
journal title
JOURNAL OF AMBIENT INTELLIGENCE AND HUMANIZED COMPUTING
volume
9
issue
2
pages
381 - 411
publisher
Springer Nature
ISSN
1868-5145
1868-5145
DOI
10.1007/s12652-017-0511-7
language
English
UGent publication?
yes
classification
U
copyright statement
I have transferred the copyright for this publication to the publisher
id
8526738
handle
http://hdl.handle.net/1854/LU-8526738
alternative location
https://doi.org/10.1007/s12652-017-0511-7
date created
2017-07-10 14:14:13
date last changed
2018-07-09 19:17:34
@article{8526738,
  abstract     = {Understanding activity patterns in office environments is important in order to increase workers{\textquoteright} comfort and productivity. This paper proposes an automated system for discovering activity patterns of multiple persons in a work environment using a network of cheap low-resolution visual sensors (900 pixels). Firstly, the users{\textquoteright} locations are obtained from a robust people tracker based on recursive maximum likelihood principles. Secondly, based on the users{\textquoteright} mobility tracks, the high density positions are found using a bivariate kernel density estimation. Then, the hotspots are detected using a confidence region estimation. Thirdly, we analyze the individual{\textquoteright}s tracks to find the starting and ending hotspots. The starting and ending hotspots form an observation sequence, where the user{\textquoteright}s presence and absence are detected using three powerful Probabilistic Graphical Models (PGMs). We describe two approaches to identify the user{\textquoteright}s status: a single model approach and a two-model mining approach. We evaluate both approaches on video sequences captured in a real work environment, where the persons{\textquoteright} daily routines are recorded over 5 months. We show how the second approach achieves a better performance than the first approach. Routines dominating the entire group{\textquoteright}s activities are identified with a methodology based on the Latent Dirichlet Allocation topic model. We also detect routines which are characteristic of persons. More specifically, we perform various analysis to determine regions with high variations, which may correspond to specific events.},
  author       = {Eldib, Mohamed and Deboeverie, Francis and Philips, Wilfried and Aghajan, Hamid},
  issn         = {1868-5145},
  journal      = {JOURNAL OF AMBIENT INTELLIGENCE AND HUMANIZED COMPUTING },
  keyword      = {Visual sensor network,Supervised learning,Probabilistic graphical models,Topic models,Sequence mining},
  language     = {eng},
  number       = {2},
  pages        = {381--411},
  publisher    = {Springer Nature},
  title        = {Discovering activity patterns in office environment using a network of low-resolution visual sensors},
  url          = {http://dx.doi.org/10.1007/s12652-017-0511-7},
  volume       = {9},
  year         = {2018},
}

Chicago
Eldib, Mohamed, Francis Deboeverie, Wilfried Philips, and Hamid Aghajan. 2018. “Discovering Activity Patterns in Office Environment Using a Network of Low-resolution Visual Sensors.” Journal of Ambient Intelligence and Humanized Computing 9 (2): 381–411.
APA
Eldib, M., Deboeverie, F., Philips, W., & Aghajan, H. (2018). Discovering activity patterns in office environment using a network of low-resolution visual sensors. JOURNAL OF AMBIENT INTELLIGENCE AND HUMANIZED COMPUTING , 9(2), 381–411.
Vancouver
1.
Eldib M, Deboeverie F, Philips W, Aghajan H. Discovering activity patterns in office environment using a network of low-resolution visual sensors. JOURNAL OF AMBIENT INTELLIGENCE AND HUMANIZED COMPUTING . Springer Nature; 2018;9(2):381–411.
MLA
Eldib, Mohamed, Francis Deboeverie, Wilfried Philips, et al. “Discovering Activity Patterns in Office Environment Using a Network of Low-resolution Visual Sensors.” JOURNAL OF AMBIENT INTELLIGENCE AND HUMANIZED COMPUTING 9.2 (2018): 381–411. Print.