Situation Awareness in Smart Home based on Multiple User Activities Recognition

• Main Authors: Lee, Ya-Hua, 李雅樺
• Other Authors: Lin, Fu-Chun
• Format: Others
• Language: en_US
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/4w4dg9

碩士 === 國立交通大學 === 資訊科學與工程研究所 === 106 === There are many research efforts on single user activity recognition in smart home to achieve situation awareness. However, a harder problem is situation awareness when multiple users are present. In this paper, we investigate the problem of recognizing multiuser activities using wearable devices in a home environment. Our research objective is to provide situation awareness so that a smart home can respond to the needs of its residents based on the accurate detection of their activities. An important part of the research is to detect whether there is any service conflict resulting from the detected activities. If so, a conflict resolution algorithm needs to be applied so that the most suitable service for the multiple users can be intelligently selected. In addition, we compare the performance of applying artificial neural network, decision tree and simple logistic regression for model construction and activity detection. Moreover, different architectural alternatives of the system design are evaluated in order to find out the best system configuration that includes wearable devices, smartphone gateway and back-end server.