Odin Severinsen
Papers
1
Total Citations
10
H-Index
1
About
Odin Severinsen is a researcher at the forefront of robotic perception and computer vision, with a primary focus on 6D object pose estimation and its integration into autonomous navigation systems. His most notable contribution is the development of SLAM-supported self-training methods, which address the critical challenge of domain shift when robots encounter unfamiliar environments. By leveraging simultaneous localization and mapping (SLAM) to generate high-quality pseudo-labels, Severinsen’s work enables object pose predictors to adapt online, significantly improving robustness and accuracy in real-world deployment. His 2022 paper, "SLAM-Supported Self-Training for 6D Object Pose Estimation," has garnered 10 citations, reflecting its early impact on the field. This work bridges the gap between geometric SLAM and semantic object understanding, offering a practical pathway for robots to build object-level scene representations during navigation. Severinsen’s research is particularly valuable for students and engineers working on embodied AI, as it tackles a fundamental bottleneck in deploying vision systems beyond controlled lab settings. His contributions are paving the way for more resilient, adaptive robots capable of operating reliably in dynamic, unstructured environments.
Research Focus
Key Achievements
Top Papers
- 1SLAM-Supported Self-Training for 6D Object Pose Estimation10 citations · 2022