A Survey of Sensor-based Activity Recognition: A Machine Learning Perspective

PhD Qualifying Examination


Title: "A Survey of Sensor-based Activity Recognition: A Machine Learning Perspective"

Mr. Hao HU


Abstract:

Automatic recognizing human activities is an important task which can support 
many applications, from context-aware computing to just-in-time information 
systems to assistive technology for the disabled. It is also a 
multidisciplinary research area draws from Machine Learning and AI, Machine 
Perception, Ubiquitous Computing, Human Computer Interaction, as well as 
psychology and sociology. Thus, human activity recognition has been drawing 
increasing interest for researchers in a variety of fields.

We focus the two major components in activity recognition systems in this 
paper, the first is the sensing component and the other is the learning and 
inference component. In this article, we would survey some of the important 
works being pursued in these two components over the past few years. We would 
also discuss some other important research topics related to activity 
recognition that is nonetheless difficult to be categorized into progresses in 
the two major components. Finally, we would also try to discuss some problems 
that exist alongside which we could carry on in other future research 
directions.


Date:     		Wednesday, 29 April 2009

Time:                   11:00am-1:00pm

Venue:                  Room 3501
 			lifts 25-26

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


**** ALL are Welcome ****