Integration and coordination in a cognitive vision system

Wrede, Sebastian, Hanheide, Marc, Wachsmuth, Sven and Sagerer, Gerhard (2006) Integration and coordination in a cognitive vision system. In: Computer Vision Systems, 2006 ICVS '06. IEEE International Conference on, 4 - 7 January 2006, New York.

Full content URL: http://dx.doi.org/10.1109/ICVS.2006.36

Documents
Wrede2006-Integration_and_Coordination_in_a_Cognitive_Vision_System_(1).pdf
[img]
[Download]
[img]
Preview
PDF
Wrede2006-Integration_and_Coordination_in_a_Cognitive_Vision_System_(1).pdf - Whole Document

611kB
Item Type:Conference or Workshop contribution (Paper)
Item Status:Live Archive

Abstract

In this paper, we present a case study that exemplifies
general ideas of system integration and coordination.
The application field of assistant technology provides an
ideal test bed for complex computer vision systems including
real-time components, human-computer interaction, dynamic
3-d environments, and information retrieval aspects.
In our scenario the user is wearing an augmented reality device
that supports her/him in everyday tasks by presenting
information that is triggered by perceptual and contextual
cues. The system integrates a wide variety of visual functions
like localization, object tracking and recognition, action
recognition, interactive object learning, etc. We show
how different kinds of system behavior are realized using
the Active Memory Infrastructure that provides the technical
basis for distributed computation and a data- and eventdriven
integration approach.

Additional Information:In this paper, we present a case study that exemplifies general ideas of system integration and coordination. The application field of assistant technology provides an ideal test bed for complex computer vision systems including real-time components, human-computer interaction, dynamic 3-d environments, and information retrieval aspects. In our scenario the user is wearing an augmented reality device that supports her/him in everyday tasks by presenting information that is triggered by perceptual and contextual cues. The system integrates a wide variety of visual functions like localization, object tracking and recognition, action recognition, interactive object learning, etc. We show how different kinds of system behavior are realized using the Active Memory Infrastructure that provides the technical basis for distributed computation and a data- and eventdriven integration approach.
Keywords:Robotics, Human-robot interaction
Subjects:H Engineering > H670 Robotics and Cybernetics
Divisions:College of Science > School of Computer Science
Related URLs:
ID Code:6943
Deposited On:30 Nov 2012 12:04

Repository Staff Only: item control page