In October 2024, Manuel Hetzel presented the results of his latest research paper at the 18th European Conference on Computer Vision (ECCV) in Milan in the workshop ‘Uncertainty Quantification for Computer Vision’. For his contribution titled ‘Reliable Probabilistic Human Trajectory Prediction for Autonomous Applications’, he was honoured with a ‘Best Paper Award’ at this workshop. Manuel Hetzel is a research assistant at the Aschaffenburg University of Applied Sciences (AUAS) Laboratory for Cooperative, Automated Transport Systems, which is headed by Prof Dr Konrad Doll.
The ECCV is held every two years by the European Computer Vision Association and is one of the world's largest and most influential scientific conferences in the booming field of computer systems and artificial intelligence. Over five days, more than 8,000 scientists from industry and research presented and discussed their latest research findings.
As part of his research work, Hetzel developed a fast, resource-friendly and reliable neural network that uses probability calculations to predict the future location of humans for the next five seconds.The method can be used in a wide range of applications, for example in the field of autonomous driving, autonomous robots and human-machine interaction.
Compared to existing methods, its methodology is characterised by its speed, the reliability achieved and the quality of the uncertainty assessment. Other authors of the paper were Hannes Reichert, Konrad Doll (all Aschaffenburg UAS) and Bernhard Sick (University of Kassel).