Unsupervised illness recognition via in-home monitoring by depth cameras

Elena Vildjiounaite (Corresponding Author), Satu-Marja Mäkelä, Tommi Keränen, Vesa Kyllönen, Ville Huotari, Sari Järvinen, Georgy Gimelfarb

Research output: Contribution to journalArticleScientificpeer-review

5 Citations (Scopus)

Abstract

Most of today's in-home systems for detecting illness in the elderly meet with the same limitations: they target only single occupancy apartments, although couples may need support, too, and, faced with monitored subjects who are not necessarily regular in their everyday life, they tend to classify even slight deviations from standard routines (e.g. cooking a meal one hour later than usual) as anomalies. The training data used for anomaly detectors typically include only days when the subject is alone and not sick, whereas in practice it is difficult to obtain day labels, i.e. information on whether the elderly subject had visitors or felt well or unwell. In addition, as every room in the apartment to be monitored is usually equipped with passive infrared motion detectors, the maintaining of such systems may be inconvenient for the residents. To address these problems, this paper proposes a new probabilistic illness detector which does not rely on regular daily routines and requires no data labelling. The detector was tested in apartments inhabited by single elderly subjects or couples, and in all cases only the living rooms and corridors were monitored, with no invasion into the more private spaces. Despite its fully unsupervised training on data covering both normal and unusual days (days of illness, visits by other people, etc.), the proposed detector distinguished between normal days and illnesses with an average accuracy of 88% and did not misclassify the receptions of guests as anomalies.
Original languageEnglish
Pages (from-to)166-187
Number of pages22
JournalPervasive and Mobile Computing
Volume38
Issue numberPart 1
DOIs
Publication statusPublished - 1 Jul 2017
MoE publication typeA1 Journal article-refereed

Fingerprint

Camera
Cameras
Detector
Monitoring
Detectors
Anomaly
Cooking
Invasion
Labeling
Labels
Infrared
Covering
Deviation
Classify
Tend
Infrared radiation
Target
Motion
Training

Keywords

  • Ambient assisted living
  • Elderly monitoring
  • Anomaly detection
  • Depth sensor-based situation awareness
  • Unsupervised learning

Cite this

@article{0e9b86eaef974c09affa53f6617c699b,
title = "Unsupervised illness recognition via in-home monitoring by depth cameras",
abstract = "Most of today's in-home systems for detecting illness in the elderly meet with the same limitations: they target only single occupancy apartments, although couples may need support, too, and, faced with monitored subjects who are not necessarily regular in their everyday life, they tend to classify even slight deviations from standard routines (e.g. cooking a meal one hour later than usual) as anomalies. The training data used for anomaly detectors typically include only days when the subject is alone and not sick, whereas in practice it is difficult to obtain day labels, i.e. information on whether the elderly subject had visitors or felt well or unwell. In addition, as every room in the apartment to be monitored is usually equipped with passive infrared motion detectors, the maintaining of such systems may be inconvenient for the residents. To address these problems, this paper proposes a new probabilistic illness detector which does not rely on regular daily routines and requires no data labelling. The detector was tested in apartments inhabited by single elderly subjects or couples, and in all cases only the living rooms and corridors were monitored, with no invasion into the more private spaces. Despite its fully unsupervised training on data covering both normal and unusual days (days of illness, visits by other people, etc.), the proposed detector distinguished between normal days and illnesses with an average accuracy of 88{\%} and did not misclassify the receptions of guests as anomalies.",
keywords = "Ambient assisted living, Elderly monitoring, Anomaly detection, Depth sensor-based situation awareness, Unsupervised learning",
author = "Elena Vildjiounaite and Satu-Marja M{\"a}kel{\"a} and Tommi Ker{\"a}nen and Vesa Kyll{\"o}nen and Ville Huotari and Sari J{\"a}rvinen and Georgy Gimelfarb",
note = "ISI: COMPUTER SCIENCE, INFORMATION SYSTEMS",
year = "2017",
month = "7",
day = "1",
doi = "10.1016/j.pmcj.2016.07.004",
language = "English",
volume = "38",
pages = "166--187",
journal = "Pervasive and Mobile Computing",
issn = "1574-1192",
publisher = "Elsevier",
number = "Part 1",

}

Unsupervised illness recognition via in-home monitoring by depth cameras. / Vildjiounaite, Elena (Corresponding Author); Mäkelä, Satu-Marja; Keränen, Tommi; Kyllönen, Vesa; Huotari, Ville; Järvinen, Sari; Gimelfarb, Georgy.

In: Pervasive and Mobile Computing, Vol. 38, No. Part 1, 01.07.2017, p. 166-187.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Unsupervised illness recognition via in-home monitoring by depth cameras

AU - Vildjiounaite, Elena

AU - Mäkelä, Satu-Marja

AU - Keränen, Tommi

AU - Kyllönen, Vesa

AU - Huotari, Ville

AU - Järvinen, Sari

AU - Gimelfarb, Georgy

N1 - ISI: COMPUTER SCIENCE, INFORMATION SYSTEMS

PY - 2017/7/1

Y1 - 2017/7/1

N2 - Most of today's in-home systems for detecting illness in the elderly meet with the same limitations: they target only single occupancy apartments, although couples may need support, too, and, faced with monitored subjects who are not necessarily regular in their everyday life, they tend to classify even slight deviations from standard routines (e.g. cooking a meal one hour later than usual) as anomalies. The training data used for anomaly detectors typically include only days when the subject is alone and not sick, whereas in practice it is difficult to obtain day labels, i.e. information on whether the elderly subject had visitors or felt well or unwell. In addition, as every room in the apartment to be monitored is usually equipped with passive infrared motion detectors, the maintaining of such systems may be inconvenient for the residents. To address these problems, this paper proposes a new probabilistic illness detector which does not rely on regular daily routines and requires no data labelling. The detector was tested in apartments inhabited by single elderly subjects or couples, and in all cases only the living rooms and corridors were monitored, with no invasion into the more private spaces. Despite its fully unsupervised training on data covering both normal and unusual days (days of illness, visits by other people, etc.), the proposed detector distinguished between normal days and illnesses with an average accuracy of 88% and did not misclassify the receptions of guests as anomalies.

AB - Most of today's in-home systems for detecting illness in the elderly meet with the same limitations: they target only single occupancy apartments, although couples may need support, too, and, faced with monitored subjects who are not necessarily regular in their everyday life, they tend to classify even slight deviations from standard routines (e.g. cooking a meal one hour later than usual) as anomalies. The training data used for anomaly detectors typically include only days when the subject is alone and not sick, whereas in practice it is difficult to obtain day labels, i.e. information on whether the elderly subject had visitors or felt well or unwell. In addition, as every room in the apartment to be monitored is usually equipped with passive infrared motion detectors, the maintaining of such systems may be inconvenient for the residents. To address these problems, this paper proposes a new probabilistic illness detector which does not rely on regular daily routines and requires no data labelling. The detector was tested in apartments inhabited by single elderly subjects or couples, and in all cases only the living rooms and corridors were monitored, with no invasion into the more private spaces. Despite its fully unsupervised training on data covering both normal and unusual days (days of illness, visits by other people, etc.), the proposed detector distinguished between normal days and illnesses with an average accuracy of 88% and did not misclassify the receptions of guests as anomalies.

KW - Ambient assisted living

KW - Elderly monitoring

KW - Anomaly detection

KW - Depth sensor-based situation awareness

KW - Unsupervised learning

UR - http://www.scopus.com/inward/record.url?scp=84995480115&partnerID=8YFLogxK

U2 - 10.1016/j.pmcj.2016.07.004

DO - 10.1016/j.pmcj.2016.07.004

M3 - Article

VL - 38

SP - 166

EP - 187

JO - Pervasive and Mobile Computing

JF - Pervasive and Mobile Computing

SN - 1574-1192

IS - Part 1

ER -