Tutorial: at AAAI 2011. August 7 - 11. San Fransisco, USA.

Recognising Behaviour in a Spatio-Temporal Context

Stacks Image 573
Tutorial Presenters:
Hans Guesgen, Massey University, New Zealand
Mehul Bhatt, University of Bremen, Germany
Stephen Marsland, Massey University, New Zealand

Download: Tutorial Syllabus
Recognising Behaviour
Recognising human behaviour plays a significant role in many applications, ranging from ensuring security in public and private places to monitoring people (e.g., with diminished or challenged mental and physical capabilities) and their interactions with systems and artefacts in a smart home, to name just two examples. Recent years have seen significant progress in methods, algorithms, and technologies for behaviour recognition, but the task remains to be a challenging one and far from being solved.
This tutorial views human behaviour recognition as the task of finding a mapping from some stream of sensor information to a sequence of recognised activities, and to classify the recognised activities according to some given classification scheme. In the smart home context for instance, the aim might be to monitor the behaviour of the inhabitant of the home and detect illness or potentially dangerous or abnormal behaviour.
About this Tutorial
The tutorial will provide an introduction to the recent developments of the area of behaviour recognition, highlighting the AI techniques that are most frequently used to achieve the task. It will argue that behaviour recognition in isolation is likely to fail, independently of which method is used. Based on this observation, it will demonstrate how considering the spatio-temporal context in which the behaviour occurs can boost the performance of the recognition process. The tutorial will conclude with highlighting the challenges and opportunities for specific sub-fields from AI research in this vibrant and emerging area of sociological, scientific and economic interest.
Prerequisites: A basic understanding of AI methods and their applications.
Duration: Half day
Goal of the Tutorial
This tutorial will introduce the audience to a topic that has an impact on many areas of AI and therefore is gaining significant importance in the community. It will offer the opportunity for AI researchers and practitioners of different areas of AI to explore the potential that behaviour recognition has for the AI community and for society in general.

Since the topic integrates many areas of AI, it is appropriate to a general audience. The tutorial will introduce the audience to applications of behaviour which are starting to have a practical impact in the real world today, but it will also familiarise the audience with the underlying methods and techniques. It will serve the following objectives:
  • Introducing the basic concepts of behaviour recognition
  • Introducing spatio-temporal representation and reasoning
    • formal methods in qualitative reasoning about space and time
    • spatio-temporal dynamics
    • reasoning about space, actions and change
  • Space, time, and the modelling of spatio-temporal context (as applicable to behaviour recognition)
  • Demonstrating the significance and role of spatio-temporal context in behaviour recognition
  • Illustrating particular applications of behaviour recognition
  • Outlining scientific challenges and future trends
Structure of the Tutorial
The tutorial will adopt the following structure:
  • Introduction
    • Motivations
    • History
    • Basic concepts
  • Current Systems
    • Areas of Application
    • Some Systems in Operation
  • AI Techniques Applied to Behaviour Recognition
    • Symbolic Approaches
    • Graphical Models
    • Spatio-Temporal Representation and Reasoning
    • Integration
  • Challenges and Future Directions
A detailed format will be made available in due course. Supporting material (slides, literature etc) for the tutorial will be posted online well ahead of the tutorial date.
Tutorial Contact
Please direct all questions and comments to Hans W. Guesgen (h.w.guesgen AT massey.ac.nz), Mehul Bhatt (bhatt AT informatik.uni-bremen.de) or Stephen Marsland (s.r.marsland AT massey.ac.nz). We value and welcome your feedback.