当前位置: X-MOL 学术J. Ambient Intell. Smart Environ. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Towards context-aware assistive applications for aging in place via real-life-proof activity detection
Journal of Ambient Intelligence and Smart Environments ( IF 1.7 ) Pub Date : 2018-11-09 , DOI: 10.3233/ais-180505
Loïc Caroux 1, 2 , Charles Consel 2, 3 , Lucile Dupuy 2 , Hélène Sauzéon 2, 4
Affiliation  

Assisted living applications can support aging in place efficiently when their context-awareness is based on a real-life-proof approach to activity detection. Recently, Caroux et al. proposed a new approach to monitoring activities dedicated to older adults, named "activity verification". This approach uses a knowledge-driven framework that draws from the literature on older adults. The purpose of the present study is to address the limitations of this approach by scaling it up and by demonstrating that it is applicable to context-aware assistive applications for aging in place. First, an experimental study was conducted in which this approach was used to monitor a large range of daily activities, for a long period (8 weeks of experimentation) and involving several participants (7 participants). Second, this approach was used to validate two examples of context-aware assisted living applications, via simulation, based on real-life sensor log data. Results showed that the applicability of the "activity veri-fication" approach scales up to a large range of daily activities by extending this approach (with accuracy values ranging between 0.82 and 1.00 depending on the activity of interest). Its inter-participant and intra-participant consistencies were demonstrated. Its limitations were addressed and the applicability to context-aware assistive applications for aging in place running on a dedicated platform was demonstrated.

中文翻译:

通过实时的活动检测来实现上下文感知的辅助应用程序,以实现适当的老化

当辅助生活应用程序的上下文感知基于对活动进行检测的真实生活方法时,它们可以有效地支持就地老化。最近,Caroux等。提出了一种新的方法来监视专门针对老年人的活动,称为“活动验证”。这种方法使用了一个知识驱动的框架,该框架借鉴了有关老年人的文献。本研究的目的是通过按比例放大并证明该方法适用于就地老化的情境感知辅助应用程序来解决该方法的局限性。首先,进行了一项实验研究,其中该方法用于监视很长一段时间(实验8周)的大范围日常活动,并涉及数名参与者(7名参与者)。第二,该方法用于基于现实生活的传感器日志数据,通过仿真来验证上下文感知辅助生活应用程序的两个示例。结果表明,“活动验证”方法的适用性通过扩展该方法而扩展到很大范围的日常活动(准确度范围在0.82到1.00之间,具体取决于感兴趣的活动)。它的参与者之间和参与者内部的一致性得到了证明。解决了它的局限性,并演示了在专用平台上运行的现场感知上下文辅助应用程序的适用性。通过扩展此方法,该方法可以扩展到各种日常活动(根据所关注的活动,其准确度范围在0.82到1.00之间)。它的参与者之间和参与者内部的一致性得到了证明。解决了它的局限性,并演示了在专用平台上运行的现场感知上下文辅助应用程序的适用性。通过扩展此方法,该方法可以扩展到各种日常活动(准确度范围在0.82到1.00之间,具体取决于感兴趣的活动)。它的参与者之间和参与者内部的一致性得到了证明。解决了它的局限性,并演示了在专用平台上运行的现场感知上下文辅助应用程序的适用性。
更新日期:2018-11-09
down
wechat
bug