Towards Real-time and Spacial-Aware Discovering of Vehicles via Acoustic Sensing

MPhil Thesis Defence


Title: "Towards Real-time and Spacial-Aware Discovering of Vehicles via 
Acoustic Sensing"

By

Mr. Hao CHEN


Abstract

Motivated by safety challenges resulting from distracted pedestrians, we 
design and implement VSpace, an acoustic-based system for oncoming vehicle 
spacial-aware discovering on smartphones via acoustic Doppler Effect, 
which aims to enhance pedestrian safety by alarming potential vehicle 
crashes. VSpace can tell the spacial status including velocity, direction 
and minimal distance of an approaching vehicle to a pedestrian online 
based on vehicle acoustic sensing. VSpace performs vehicle spacial-aware 
discovering in four major steps: (1)Vehicle acoustic signal processing; 
(2)Vehicle classification; (3)Vehicle sound Doppler profiling and 
(4)Vehicle spatial status reconstruction. All implementation of VSpace is 
based on COTS smartphone with microphone. We evaluate VSpace with 
comprehensive experiments of 20 different vehicles in 240 practical cases, 
in which VSpace achieves an overall accuracy of 91.67% in vehicle 
classification and satisfactory low relative error in vehicle spacial 
status reconstruction.


Date:			Wednesday, 20 September 2017

Time:			4:30pm - 6:30pm

Venue:			Room 5566
 			Lifts 27/18

Committee Members:	Dr. Ke Yi (Supervisor)
 			Dr. Kai Chen (Chairperson)
 			Prof. Gary Chan


**** ALL are Welcome ****