Abstract
With the global population ageing, it is crucial to enable
individuals to live independently and safely in their homes. Using ubiq-
uitous sensors such as Passive InfraRed sensors (PIR) and door sensors is
drawing increasing interest for monitoring daily activities and facilitat-
ing preventative healthcare interventions for the elderly. Human Activ-
ity Recognition (HAR) from passive sensors mostly relies on traditional
machine learning and includes data segmentation, feature extraction,
and classification. While techniques like Sensor Weighting Mutual Infor-
mation (SWMI) capture spatial context in a feature vector, effectively
leveraging temporal information remains a challenge. We tackle this by
clustering activities into morning, afternoon, and night, and encoding
them into the feature weighting method calculating distinct mutual in-
formation matrices. We further propose to extend the feature vector by
incorporating time of day and day of week as cyclical temporal features,
as well as adding a feature to track the user’s location. The experiments
show improved accuracy and F1-score over existing state-of-the-art meth-
ods in three out of four real-world datasets, with highest gains in a low-
data regime. These results highlight the potential of our approach for
developing effective smart home solutions to support ageing in place.
individuals to live independently and safely in their homes. Using ubiq-
uitous sensors such as Passive InfraRed sensors (PIR) and door sensors is
drawing increasing interest for monitoring daily activities and facilitat-
ing preventative healthcare interventions for the elderly. Human Activ-
ity Recognition (HAR) from passive sensors mostly relies on traditional
machine learning and includes data segmentation, feature extraction,
and classification. While techniques like Sensor Weighting Mutual Infor-
mation (SWMI) capture spatial context in a feature vector, effectively
leveraging temporal information remains a challenge. We tackle this by
clustering activities into morning, afternoon, and night, and encoding
them into the feature weighting method calculating distinct mutual in-
formation matrices. We further propose to extend the feature vector by
incorporating time of day and day of week as cyclical temporal features,
as well as adding a feature to track the user’s location. The experiments
show improved accuracy and F1-score over existing state-of-the-art meth-
ods in three out of four real-world datasets, with highest gains in a low-
data regime. These results highlight the potential of our approach for
developing effective smart home solutions to support ageing in place.
Original language | English |
---|---|
Title of host publication | Proceedings of the International Conference on Ubiquitous Computing and Ambient Intelligence (UCAmI 2024) |
Editors | José Bravo, Chris Nugent, Ian Cleland |
Publisher | Springer |
Pages | 238-251 |
Number of pages | 14 |
ISBN (Electronic) | 9783031775710 |
ISBN (Print) | 9783031775703 |
DOIs | |
Publication status | Published - 21 Dec 2024 |
Publication series
Name | Lecture Notes in Networks and Systems |
---|---|
Volume | 1212 LNNS |
ISSN (Print) | 2367-3370 |
ISSN (Electronic) | 2367-3389 |
Bibliographical note
Copyright © Springer Nature B.V. 2024. This version of the article has been accepted for publication, after peer review and is subject to Springer Nature’s AM terms of use but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: https://link.springer.com/chapter/10.1007/978-3-031-77571-0_24https://link.springer.com/book/9783031775703Keywords
- Activity Recognition
- Data Segmentation
- Mutual Information
- Sliding Window
- Smart Home
- Temporal Context