Implicit Mobile Crowdsourcing for Fingerprint Database Construction: Approaches and Comparisons

PhD Qualifying Examination


Title: "Implicit Mobile Crowdsourcing for Fingerprint Database Construction: 
Approaches and Comparisons"

by

Mr. Weipeng ZHUO


Abstract:

Indoor localization is becoming more and more important in people's daily life. 
However, construction of fingerprint database for localization is still labor 
intensive and time consuming. Among various techniques, implicit mobile 
crowdsourcing has emerged as a practical and popular approach. It is cost 
effective, non-intrusive to users and easily scalable. In this survey, we study 
from the literature important works on implicit mobile crowdsourcing for 
fingerprint database construction, to provide a comprehensive overview of this 
field. In particular, we focus on two areas of research, pure sensor signal 
based approach and fusion of different signals. Pure sensor signal based 
techniques leverage single signal from mobile devices to conduct crowdsourcing 
behavior and label the crowdsourced data accordingly. Commonly used signals 
include inertial navigation sensors (INS), radio frequencies (RF), sound and 
light, etc. Fusion-based approach takes advantage of characteristics of each 
signal so as to make a better prediction for labels of implicit crowdsourced 
signals. Fusion of INS and RF plays an important role in implicit 
crowdsourcing, while others are also evolving quickly to tackle the problem. In 
this survey, we review extensively both classic and recent advances for the two 
categories of schemes. We also study and compare their strengths and weaknesses 
under practical deployment.


Date:			Friday, 30 August 2019

Time:                  	10:00am - 12:00noon

Venue:                  Room 3494
                         Lifts 25/26

Committee Members:	Prof. Gary Chan (Supervisor)
 			Dr. Brian Mak (Chairperson)
 			Prof. Andrew Horner
 			Prof. Ke Yi


**** ALL are Welcome ****