MobVis

Vision Technologies and Intelligent Maps for Mobile Attentive Interfaces in Urban Scenarios

Project duration: May 1st, 2005 to April 30th, 2008

The main objective in MOBVIS is to achieve a theoretical and practical leap in the application of artificial vision in smart mobile applications with a primary focus in spatial awareness and guidance. In order to achieve this goal, MOBVIS concentrates its research on the integration of multi-modal context awareness, vision based object recognition, and intelligent map technology, into an innovative form of an attentive interface, which enables perception and reasoning on a vast amount of data and in a continuously operating framework.

Mobile Vision in Urban Scenarios

Partners:

Joanneum research, Graz, Austria
Royal Institute of Technology (KTH), Stockholm, Sweden
Technical University of Darmstadt, Germany
TeleAtlas, Gent, Belgium

Our scientific contributions lie within the following topics:

Informative Visual Features
Image Based Localization
Recognition of Parts, Objects and Events
Vision Based Context
Incremental Updating of the Intelligent Map
High-level Map Indexing and Retrieval

In the course of the project we have collected the following data sets:

Scientific output of our work within the project is described in these publications:

U. Steinhoff, D. Omerčević, R. Perko, B. Schiele, and A. Leonardis, “How Computer Vision Can Help in Outdoor Positioning”, Ambient Intelligence 2007, pages 124-141, Darmstadt, Germany, 2007, (pdf).
D. Omerčević, O. Drbohlav, and A. Leonardis, “High-Dimensional Feature Matching: Employing the Concept of Meaningful Nearest Neighbors”, ICCV 2007, Rio de Janeiro, Brazil, 2007, (pdf).
R. Perko and A. Leonardis. Attention in Cognitive Systems. Theories and Systems from an Interdisciplinary Viewpoint (WAPCV 2007), volume 4840, chapter “Context Driven Focus of Attention for Object Detection”, pages 216-233. Springer LNAI, December 2007, (pdf).