VAMPIRE___Visual Active Memory Processes and Interactive REtrieval
VAMPIRE Events Publications Consortium Media archive
Pommersfelden 2005 Pommersfelden 2004 IST Event 2004

2nd Joint VAMPIRE Workshop with Industry

June 29/30, 2005, Pommersfelden near Erlangen, Germany

The Pommersfelden Workshops aim at establishing a dialog between Cognitive Computer Vision research as conducted by the VAMPIRE consortium and industry experience, applications and demands.

The first workshop in this series successfully started this dialog and presented work centered around the VAMPIRE project. This second workshop is more focused on specific Cognitive Vision topics that are related to work done in VAMPIRE.



Technical Program

Wed 29th of June, 13:00-19:15

  • Opening of Workshop and Overview of the VAMPIRE project

Session 1: Human-Machine Interaction and Evaluation (13:30-15:00)

  • Christoph Gräßl: Comparative Evaluation of State-of-the-Art Object Tracking Techniques
  • Gernot A. Fink: A Multi-modal Attention System for Robotic Assistents
  • Christoph Stock: Tracking by Speed for 3D Human Computer Interaction
  • Sven Wachsmuth: A System for Cognitive Assistence: First Evaluation Experiments

Demonstrations by the VAMPIRE project (15:00-17:30)

Session 2: Perspectives and Applications of Cognitive Vision (17:30-19:15)

  • Don Braggins: 'milestones' in the development of the commercially available technology in computer vision
  • Andreas Kröpfl: TriCam and its Application to Traffic Telematics
  • Gisbert Lawitzky: Applications of Cognitive Vision in Robotics
  • Matthias Hackel: face to face interaction applications with humanoid robots
  • Wolfgang Reitberger: Cognitive Vision for Natural Interaction in Ambient Environments

Thu 30th of June, 9:00-12:30

Session 3: Memory and Representation (9:00-10:20)

  • Josef Kittler: A memory architecture and contextual reasoning framework for cognitive vision
  • Jan-Olof Eklundh: Objects, categories and the role of 3D cues
  • Hans Burkhardt: Thinking in categories and equivalence classes - new results in content based image retrieval
  • Stefan Posch: A mosaic-based iconic memory with applications for a mobile robot

Session 4: Context and Learning (10:50-12:10)

  • Christian Wöhler: Autonomous self-learning of classification modules in real-time vision systems
  • Markus Vincze: ActIPret: contextual control for activity interpretation
  • Edgar Körner: Online learning of 3D objects and real-time whole-body movement control: Paving the way for learning by exploration
  • Helge Ritter: Interactive online learning in an AR scenario

Session 5: Final discussion (12:10-12:30)