Recognizing Human Activities from Physical and Virtual Worlds

PhD Thesis Proposal Defence


Title: "Recognizing Human Activities from Physical and Virtual Worlds"

by

Mr. Hao HU


ABSTRACT:

Recognizing the activities, goals and intentions underlying human actions 
has become an extensive research topic since early 1980s. Although many 
new technologies and methods have been proposed to tackle activity 
recognition problems, deploying activity recognition systems in the real 
world still encounters many practical problems. In this proposal, we aim 
to study problems regarding recognizing human activities in the physical 
world and also discuss some possible extensions of bridging the 
information gap between sensor data in the physical world and social 
activities in the virtual world. We start by surveying related works in 
these two areas followed by a study of several challenging problems in the 
area of sensor-based activity recognition. We also discuss one scenario 
which helps inferring human demographics in the physical world by using 
the data of online social check-in activities in the virtual world. 
Preliminary results of these problems are also presented and discussed.


Date:                   Wednesday, 19 September 2012

Time:                   10:00am - 12:00noon

Venue:                  Room 3501
                         lifts 25/26

Committee Members:      Prof. Qiang Yang (Supervisor)
                         Dr. Raymond Wong (Chairperson)
 			Prof. Dit-Yan Yeung
 			Dr. Ke Yi


**** ALL are Welcome ****