An Unsupervised Learning-Based Spatial Co-Location Detection System from Low-Power Consumption Sensor.

David Ishak Kosasih,Mohammed Atiquzzaman,Byung-Gook Lee,Hyotaek Lim

doi:10.3390/s21144773

David Ishak Kosasih, Mohammed Atiquzzaman + Show 2 more

Open Access

https://doi.org/10.3390/s21144773

Copy DOI

Abstract

Spatial co-location detection is the task of inferring the co-location of two or more objects in the geographic space. Mobile devices, especially a smartphone, are commonly employed to accomplish this task with the human object. Previous work focused on analyzing mobile GPS data to accomplish this task. While this approach may guarantee high accuracy from the perspective of the data, it is considered inefficient since knowing the object’s absolute geographic location is not required to accomplish this task. This work proposed the implementation of the unsupervised learning-based algorithm, namely convolutional autoencoder, to infer the co-location of people from a low-power consumption sensor data—magnetometer readings. The idea is that if the trained model can also reconstruct the other data with the structural similarity (SSIM) index being above 0.5, we can then conclude that the observed individuals were co-located. The evaluation of our system has indicated that the proposed approach could recognize the spatial co-location of people from magnetometer readings.

Highlights

A spatial co-location can be defined as a set of objects which co-exist in close geographic proximity [1]
A spatial co-location detection system could be the key to control the spreading of infectious disease during an epidemic, or pandemic situation [11,12]
We propose an unsupervised learning-based system that uses mobile magnetometer readings to infer the spatial co-location of people, similar to that of Kosasih et al [8]

Summary

Introduction

A spatial co-location can be defined as a set of objects which co-exist in close geographic proximity [1]. Spatial co-location detection refers to the task of inferring the co-location of two or more objects in the geographic space [2,3]. There are many sensors that can be deployed to accomplish this task, with global positioning systems (GPS) being the commonly used sensor [13,14] Mobile devices, such as smartphones and smartwatches, have emerged as one of the most common ways to perform location tracking-related tasks, including spatial co-location detection, with the human object. This is because the majority of people already carry these devices with them everywhere.

Methods

Results

Conclusion