Igor Gilitschenski

About

Igor Gilitschenski is a leading researcher in robotics and computer vision, whose work centers on robust estimation, visual-inertial mapping, and long-term autonomous navigation. His most influential contribution is the development of **maplab**, an open-source framework for visual-inertial mapping and localization that has garnered over 270 citations and become a cornerstone for research in robust, drift-free pose estimation. Gilitschenski has also pioneered techniques for **TSDF-based change detection**, enabling robots to maintain consistent dense 3D reconstructions over extended periods by identifying dynamic objects and environmental changes. His work on **robust estimation** addresses fundamental challenges in pose estimation, point cloud alignment, and object tracking, providing algorithms resilient to outliers and sensor noise. Additionally, he has advanced **efficient descriptor learning** for large-scale localization and **appearance-based landmark selection** to improve visual localization under varying conditions. His practical innovations include **ShadowCam**, a real-time system for detecting moving obstacles behind corners, critical for autonomous vehicle safety. With over 500 cumulative citations across his top papers, Gilitschenski’s research bridges theoretical estimation problems with deployable solutions, making him a key figure in enabling reliable, long-term autonomy for robots and autonomous systems.

Research Focus

Key Achievements

Top Papers

1. 1
   Maplab: An Open Framework for Research in Visual-Inertial Mapping and Localization

   270 citations · 2018
2. 2
   TSDF-based change detection for consistent long-term dense reconstruction and dynamic object discovery

   68 citations · 2017
3. 3
   Evaluation of Combined Time-Offset Estimation and Hand-Eye Calibration on Robotic Datasets

   43 citations · 2017
4. 4
   Robust Estimation and Applications in Robotics

   38 citations · 2016
5. 5
   Robust Estimation and Applications in Robotics

   28 citations · 2016
6. 6
   Efficient descriptor learning for large scale localization

   27 citations · 2017
7. 7
   Erasing bad memories: Agent-side summarization for long-term mapping

   24 citations · 2016
8. 8
   Appearance‐based landmark selection for visual localization

   18 citations · 2019
9. 9
   Onboard real-time dense reconstruction of large-scale environments for UAV

   16 citations · 2017
10. 10
    ShadowCam: Real-Time Detection of Moving Obstacles Behind A Corner For Autonomous Vehicles

    15 citations · 2018

Key Collaborators

Contact & Links