English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

People living with dementia often exhibit behavioural and psychologicalsymptoms of dementia that can put their and others' safety at risk. Existingvideo surveillance systems in long-term care facilities can be used to monitorsuch behaviours of risk to alert the staff to prevent potential injuries ordeath in some cases. However, these behaviours of risk events are heterogeneousand infrequent in comparison to normal events. Moreover, analyzing raw videoscan also raise privacy concerns. In this paper, we present two novelprivacy-protecting video-based anomaly detection approaches to detectbehaviours of risks in people with dementia. We either extracted body poseinformation as skeletons or used semantic segmentation masks to replacemultiple humans in the scene with their semantic boundaries. Our work differsfrom most existing approaches for video anomaly detection that focus onappearance-based features, which can put the privacy of a person at risk and isalso susceptible to pixel-based noise, including illumination and viewingdirection. We used anonymized videos of normal activities to train customizedspatio-temporal convolutional autoencoders and identify behaviours of risk asanomalies. We showed our results on a real-world study conducted in a dementiacare unit with patients with dementia, containing approximately 21 hours ofnormal activities data for training and 9 hours of data containing normal andbehaviours of risk events for testing. We compared our approaches with theoriginal RGB videos and obtained a similar area under the receiver operatingcharacteristic curve performance of 0.807 for the skeleton-based approach and0.823 for the segmentation mask-based approach.

Privacy-Protecting Behaviours of Risk Detection in People with Dementia using Videos