| Title |
Towards Accurate And Robust Fall Detection For The Elderly In A Hybrid Cloud-Edge Architecture |
| ID_Doc |
57972 |
| Authors |
Shahiduzzman K.M.; Peng J.; Gao Y.; Hei X.; Cheng W. |
| Year |
2019 |
| Published |
Proceedings - 2019 IEEE SmartWorld, Ubiquitous Intelligence and Computing, Advanced and Trusted Computing, Scalable Computing and Communications, Internet of People and Smart City Innovation, SmartWorld/UIC/ATC/SCALCOM/IOP/SCI 2019 |
| DOI |
http://dx.doi.org/10.1109/SmartWorld-UIC-ATC-SCALCOM-IOP-SCI.2019.00290 |
| Abstract |
Falls are a very important phenomenon for elderly citizens because these can create fatal and non-fatal injuries and an effective fall detection technique is needed to prevent this hazard. In this paper, we are going to introduce a hybrid cloud-edge based architecture for elderly fall detection. This architecture consists of four tiers which are body area network (BAN), intelligent mobile edge, secure network and medical cloud. The BAN is composed of a number of sensors like accelerometer, gyroscope, camera etc. which can send valuable information about human activities. Data fusion of these sensors can be very useful in case of elderly fall detection in terms of achieving higher accuracy and lower false alarm rates (FAR)s. So here, we are going to introduce an algorithm, named, ActDec-SysOpt (Activity Detection System Optimization) which is based on long short term memory (LSTM) type machine learning technique focusing on fusion of accelerometer and gyroscope data to determine human activities particularly. The result shows that maximum 91 % accuracy with minimum 1.89 % FAR in detecting human activities can be achievable using these proposed algorithms. © 2019 IEEE. |
| Author Keywords |
Cloud-Network-Edge; Fall detection; Human activities; Inertia sensors; LSTM |
