Cognitive Computer Vision Group

This page is outdated and will no longer be maintained.
Since March 2016, I am at the University of Hamburg. You can find my new homepage here.

PD Dr. Simone Frintrop

Privatdozentin (Assistant professor)

Postal address: Institut für Informatik III (Institute of Computer Science III),
Rheinische Friedrich-Wilhelms-Universität Bonn
Römerstr. 164
D-53117 Bonn, Germany
Office: Room A 120
Phone: +49 (0) 228 73 4357


  • I am heading the Cognitive Computer Vision Group at the Institute of Computer Science III at the university of Bonn.
  • In our research on cognitive computer vision, we investigate concepts of the human visual system and develop algorithms that follow these mechanisms to solve computer vision tasks. We follow this approach since we believe that there are great opportunities for technical systems in better understanding and exploiting the concepts that make human vision so powerful. Especially the generalization capabilities of human perception as well as the ability to cope with noise, clutter, and with new situations still outperform current machine vision systems considerably. Vision systems of the future have to be robust, flexible, efficient, and intuitive to interact with — properties that humans have and that are worth being investigated in depth.

    The research topics of our group include object discovery, visual attention and saliency detection, visual tracking of objects and persons, and visual SLAM (simultaneous localization and mapping). Find more about our research on my research page.

    Additionally, have organized the Bonn Vision Workshop 2009 and its successor, the Bonn Vision Workshop 2013, and I was an editor of the German Journal on Artificial Intelligence (KI-Zeitschrift) from 2010 - 2014. The issue about "Bio-inspired Vision Systems" (appeared Feb. 2015, guest editors: Michael Zillich and Norbert Krüger), which I supervised as a liason editor (Pate), can be found here.


    1) We obtained the best paper award for our ICVS 2015 paper: "Saliency-Guided Object Candidates Based on Gestalt Principles", Thomas Werner, Germán Martín García, and Simone Frintrop.

    2) The code of our VOCUS2 saliency system (CVPR 2015) is online.