Matthias Rottmann
Papers
2
Total Citations
4
H-Index
2
About
Matthias Rottmann is a leading researcher in safe autonomous perception, specializing in uncertainty quantification, anomaly detection, and object recognition for self-driving cars and robotics. His work addresses a critical challenge: ensuring perception systems can reliably identify unknown or out-of-distribution objects—like wild animals or unusual debris—that were absent from training data. Rottmann’s major contributions include developing lightweight prediction quality estimators for object detection in LiDAR point clouds, as demonstrated in his 2024 paper "LMD," which enables real-time, resource-efficient assessment of detection reliability. He also co-led the creation of the OoDIS benchmark (2025), a comprehensive evaluation framework for anomaly instance segmentation and detection, providing a standardized testbed for advancing safe navigation. Though his most-cited works are recent, with 2 citations each, their impact is growing rapidly within the autonomous systems community. Rottmann’s research bridges the gap between theoretical uncertainty modeling and practical deployment, making him a key figure in the push toward robust, fail-safe perception for self-driving cars and robots.
Research Focus
Key Achievements
Top Papers
- 1
- 2OoDIS: Anomaly Instance Segmentation and Detection Benchmark2 citations · 2025