Earlier this month, PoDIUM was present at the 28th International Conference on Information Fusion (FUSION 2025), held in Rio de Janeiro, Brazil, from 7 to 11 July. It was the first time the event took place in South America, bringing together experts from academia, industry and the public sector to explore the latest developments in information fusion technologies, from multimodal sensor fusion to machine learning and human intelligence.
Alexander Scheible, from PoDIUM partner Ulm University, presented a paper titled “Online Monitoring for Multi-Object Measurement Model Parameters“. The paper focuses on methods to monitor key parameters in multi-object tracking systems used in the PoDIUM German Living Lab. It examines how variations in detection probability and clutter intensity affect filter performance, and explores methods for identifying these variations in real-time during system execution without requiring ground-truth data. The results show that the monitoring system can detect changes in both parameters and can therefore be used to improve tracking performance.
This work contributes to PoDIUM’s efforts to enhance the reliability of connected and automated mobility systems.
The presentation can be downloaded here. The full paper will be available soon on the PoDIUM website – stay tuned!