Contrastive Analysis for Human Activity Recognition Algorithms Using WiFi Signals

DOI

10.2991/icwcsn-16.2017.65How to use a DOI?

Keywords

channel state information (CSI); dynamic time warping (DTW); earth mover distance (EMD); activity recognition

Abstract

Human Activity monitoring has become increasingly important and has the potential to support a wide area of applications including elder care, well-being management, fitness tracking and building surveillance. Traditional approaches involve wearable sensors and specialized hardware installations. Compared with these solutions, channel state information (CSI) has its advantage. The algorithm for human detection based on CSI Info has become increasingly important. Some prior WiFi signal based human activity recognition systems have been proposed such as Wisee[8], WiFall[11], Witrack[9], CARM[10]. Different from prior works, we propose a contrastive analysis for the recognition algorithms under different transmission frequencies and activities. Finally, the experimental performance of DTW (Dynamic Time Wrapping) and EMD (Earth Mover Distance) is adopted. Our works show that EMD has a better performance than DTW in most cases.

Copyright

© 2017, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Jian Zhou
AU  - Han Su
AU  - Kai Yu
PY  - 2016/12
DA  - 2016/12
TI  - Contrastive Analysis for Human Activity Recognition Algorithms Using WiFi Signals
BT  - Proceedings of the 3rd International Conference on Wireless Communication and Sensor Networks (WCSN 2016)
PB  - Atlantis Press
SP  - 299
EP  - 303
SN  - 2352-538X
UR  - https://doi.org/10.2991/icwcsn-16.2017.65
DO  - 10.2991/icwcsn-16.2017.65
ID  - Zhou2016/12
ER  -