Dermot Kerr
Papers
19
Total Citations
158
H-Index
7
About
Dermot Kerr is a robotics and computer vision researcher whose work spans simultaneous localization and mapping (SLAM), autonomous driving perception, and intelligent robotic systems. He is perhaps best known for his pioneering contributions to object-level SLAM using quadric landmark representations, a body of work that has significantly advanced how robots and autonomous vehicles perceive and model their environments. His landmark 2021 paper, "Accurate and Robust Object SLAM With 3D Quadric Landmark Reconstruction in Outdoors," has garnered 35 citations and laid the foundation for a productive research thread exploring quadric initialization, dynamic scene handling, and semantic mapping — themes he has continued to develop through 2024 with the DynaQuadric framework. Beyond SLAM, Kerr has made meaningful contributions to 3D object detection using LiDAR data, event-camera-based robotic vision, meta-reinforcement learning for manipulation, and early work in automated code generation for mobile robots. His 2008 study comparing cornerness measures for interest point detection reflects a career-long commitment to robust feature extraction. Collectively, his publications demonstrate a sustained and evolving research vision aimed at enabling robots to perceive, navigate, and interact intelligently with complex real-world environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5BSH-Det3D: Improving 3D Object Detection with BEV Shape Heatmap9 citations · 2023
- 6Towards Automated Code Generation for Autonomous Mobile Robots9 citations · 2010
- 7
- 8
- 9Comparing Cornerness Measures for Interest Point Detection7 citations · 2008
- 10Biologically inspired intensity and range image feature extraction7 citations · 2013