http://www.dagstuhl.de/10251

June 20 – 25 , 2010, Event 10251

Probabilistic Methods for Perceiving, Learning and Reasoning about Everyday Activities

Organizers

Michael Beetz (TU München, DE)
Tanzeem Choudhury (Dartmouth College – Hanover, US)
Andreas Krause (ETH Zürich, CH)
Anna Schubö (LMU München, DE)

For support, please contact

Heike Clemens

Participants and Materials

http://www.dagstuhl.de/Materials/10251

Motivation

The Dagstuhl event “Probabilistic Methods for Perceiving, Learning and Reasoning about Everyday Activities” addresses a key technology for Ambient Assistive Living, where the goal is to achieve the prolonged independence of elderly and partially disabled people. This is a particularly pressing aspect of an aging society. Within this research field, the acquisition, learning and use of probabilistic models of everyday activities presents a key interdisciplinary challenge for researchers in artificial intelligence (in particular in probabilistic reasoning, machine learning, image understanding and knowledge representation), cognitive science and psychology, health care and robotics. Progress towards this challenge will have an enormous impact on

  • understanding the relationship between health, diseases, and motion;
  • assistive cognition and ambient assistive living;
  • building models of actions and movement control in cognitive psychology, sports, biomechanics and cognitive neuroscience;
  • commonsense knowledge for personal robot assistants; and
  • smarter computing environments.

The event is to establish an interdisciplinary research community by bringing together researchers from various fields that are all equally important to the challenging research endeavor of accurately modeling everyday activities.

The goal of the Dagstuhl event is to foster new research collaborations in the context of the general task of understanding human activities and applying learnt models of everyday activities in various ways.

The concrete goals are

  • the acquisition of appropriately represented observations of human activities that are suitable for informative analysis,
  • the learning of abstract, general, semantically grounded, probabilistic models of human activities which support accurate and scalable reasoning mechanisms,
  • the application of these models to human assistance systems, such as the medically inspired applications outlined above,
  • the transfer of the acquired knowledge about everyday activities to cognitive technical systems such as personal robots and smart computing environments.

The event will include short introductory position/research statements by each participant, a small number of overview talks on the respective research fields, working group discussions and plenary sessions.

Classification

  • Artificial Intelligence / Machine Learning / Probabilistic Reasoning / Robotics / Cognition

Keywords

  • Probabilistic models
  • Probabilistic reasoning
  • Sensing
  • Cognition
  • Control
  • Learning

Online Publications

We offer several possibilities to publish the results of your event. Please contact marc.herbstritt(at)dagstuhl.de if you are interested.

Dagstuhl's Impact

Please inform us when a publication was published as a result from your seminar. These publications are listed in the category Dagstuhl's Impact and are presented on a special shelf in the library.