Papers

12

Total Citations

840

H-Index

8

About

Benedikt Mersch is a leading researcher in robotics perception, specializing in 3D LiDAR-based odometry, mapping, and dynamic scene understanding. His most impactful contribution is **KISS-ICP**, a deceptively simple yet remarkably robust point-to-point ICP registration framework that challenges the trend toward algorithmic complexity in sensor odometry. With over 435 citations, this work has become a gold standard for accurate and reliable ego-motion estimation. Mersch also pioneered learning-based **moving object segmentation** in 3D LiDAR data (237 citations), enabling robots to distinguish static from dynamic elements in real time—a critical capability for safe navigation and consistent map building. He extended this work into **ERASOR2** for instance-aware static mapping and **Radar Instance Transformer** for moving instance segmentation under adverse weather, demonstrating cross-sensor expertise. His recent contributions include **Kinematic-ICP**, which incorporates wheeled robot constraints for planar motion, and self-supervised geometric scan completion for dense mapping. Mersch’s work consistently emphasizes practical, deployable solutions that balance accuracy with computational efficiency, making him a key figure in advancing robust autonomy for mobile robots operating in dynamic, real-world environments.

Research Focus

Key Achievements

8
H-Index
12
Papers
840
Total Citations
70
Avg Citations/Paper
🏆 Most Cited Paper
KISS-ICP: In Defense of Point-to-Point ICP – Simple, Accurate, and Robust Registration If Done the Right Way
435 citations · 2023
📈 Most Prolific Year: 2023 (4 Papers)
🤝 Key Collaborators: 26
🏛 Institutions: University of Bonn, Hamburg University of Technology, Robotics Research (United States)

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 13 days ago