• Home
• What is IK?
• Focus Theme 2011
• Organization
• Program
  • IK 2011 Schedule
• Poster and Demo Session
• Location
• Stipends
• Registration
• Contact

Speaker:

Sebastian Bader

Summary:

The course covers the basics of automated activity and intention recognition.
As a self-contained introduction it is suited for people with different
background. The course shall give an overview of existing approaches and
present the underlying ideas rather than all technical details.

After motivating the topic, we discuss a successful approach to automatic
activity recognition based on mobile sensors. In this first part algorithms
known from machine learning are introduced and combined into a running system. If time permits, small experiments with the audience are conducted to show the system in action. All concepts are presented on a very general level and thus also suited for people without background on the subject.

Based on recognised primitive activities and different forms of available
background knowledge, we then discuss how to infer the current intentions of
the user or of a group of users. Here we employ probabilistic models, which
again are introduced on a general and intuitive level.

To wrap up, we discuss possible extensions and application domains for both,
automated activity and intention recognition.

Disciplines:
- computer science
- machine learning
- non-standard human / computer interaction

References:

• Russel, Norvig: "AI a modern approach"

CV:
Sebastian Bader studied computer science and computational logic at the Technische Universität Dresden, Germany, where he also completed his PhD. The topic of his PhD-thesis was Neural-Symbolic Integration. Since 2008 he works at the university of Rostock, Germany. His current research interest include automated intention recognition, strategy synthesis to support users in instrumented environments, and the design and control of such environments.