Probabilistic Interpretation of Situations and Behavior Recognition

The subject of this research interest is the analysis and development of robust and efficient methods for state estimation and prediction of traffic situations as the basis of autonomous behavior decisions. Due to the incomplete and uncertain observable environment, the dealing with uncertainties playes a crucial role in the fusion and estimation process. On the basis of machine learning techniques, the vehicle will be able to learn to assess situations and future developments automatically from its observations to enable safe and anticipatory driving.