Olaf Booij

About

Olaf Booij is a robotics and computer vision researcher whose work centers on visual localization, appearance-based mapping, and human-robot interaction. His most influential contributions address how mobile robots can navigate and understand their environments using visual information alone, without relying on traditional metric maps. His 2007 paper on appearance-based topological mapping (131 citations) remains a foundational reference in the field, demonstrating how robots can build rich spatial representations from camera imagery to support reliable localization and navigation. Complementary work on sparse modeling and image-set pruning improved the computational efficiency of these systems, making them more practical for real-world deployment. Booij also made meaningful strides in socially situated robotics, contributing to the BIRON project (95 citations), which explored how robots could learn new places in home environments through natural spoken dialogue — a critical capability for elderly care applications. His earlier work on translating raw sensor data into human-interpretable spatial concepts further bridged the gap between machine perception and semantic communication. More recently, Booij has turned his attention to cross-view metric localization (34 citations), developing methods that match ground-level camera images against aerial or satellite imagery — a promising direction for autonomous vehicle self-localization. Across more than two decades of research, his work consistently advances the goal of robots that perceive, reason about, and communicate spatial knowledge the way humans do.

Research Focus

Key Achievements

Top Papers

1. 1
   Navigation using an appearance based topological map

   131 citations · 2007
2. 2
   From images to rooms

   96 citations · 2006
3. 3
   BIRON, where are you? Enabling a robot to learn new places in a real home environment by integrating spoken dialog and visual localization

   95 citations · 2006
4. 4
   Visual Cross-View Metric Localization with Dense Uncertainty Estimates

   34 citations · 2022
5. 5
   From Sensors to Human Spatial Concepts: An Annotated Data Set

   24 citations · 2008
6. 6
   Cross-View Matching for Vehicle Localization by Learning Geographically Local Representations

   21 citations · 2021
7. 7
   Sparse appearance based modeling for robot localization

   21 citations · 2006
8. 8
   A geometrically constrained image similarity measure for visual mapping, localization and navigation

   16 citations · 2007
9. 9
   The planar two point algorithm

   8 citations · 2009
10. 10
    Pruning the image set for appearance based robot localization

    8 citations · 2005

Key Collaborators

Contact & Links