Difference between revisions of "Publications:Efficient Activity Recognition in Smart Homes Using Delayed Fuzzy Temporal Windows on Binary Sensors"

From ISLAB/CAISR
(Created page with "<div style='display: none'> == Do not edit this section == </div> {{PublicationSetupTemplate|Author=Rebeen Ali Hamad, Alberto Salguero Hidalgo, Rafik Bouguelia, Macarena Espin...")
 
(No difference)

Latest revision as of 22:22, 11 February 2020

Do not edit this section

Keep all hand-made modifications below

Title Efficient Activity Recognition in Smart Homes Using Delayed Fuzzy Temporal Windows on Binary Sensors
Author Rebeen Ali Hamad and Alberto Salguero Hidalgo and Mohamed-Rafik Bouguelia and Macarena Espinilla Estevez and Javier Medina Quero
Year 2020
PublicationType Journal Paper
Journal IEEE journal of biomedical and health informatics
HostPublication
Conference
DOI http://dx.doi.org/10.1109/JBHI.2019.2918412
Diva url http://hh.diva-portal.org/smash/record.jsf?searchId=1&pid=diva2:1392777
Abstract Human activity recognition has become an activeresearch field over the past few years due to its wide applicationin various fields such as health-care, smart homemonitoring, and surveillance. Existing approaches for activityrecognition in smart homes have achieved promisingresults. Most of these approaches evaluate real-timerecognition of activities using only sensor activations thatprecede the evaluation time (where the decision is made).However, in several critical situations, such as diagnosingpeople with dementia, “preceding sensor activations”are not always sufficient to accurately recognize theinhabitant’s daily activities in each evaluated time. Toimprove performance, we propose a method that delaysthe recognition process in order to include some sensoractivations that occur after the point in time where thedecision needs to be made. For this, the proposed methoduses multiple incremental fuzzy temporal windows toextract features from both preceding and some oncomingsensor activations. The proposed method is evaluated withtwo temporal deep learning models (convolutional neuralnetwork and long short-term memory), on a binary sensordataset of real daily living activities. The experimentalevaluation shows that the proposed method achievessignificantly better results than the real-time approach,and that the representation with fuzzy temporal windowsenhances performance within deep learning models. © Copyright 2020 IEEE