Feature Relevance Estimation for Learning Pedestrian Behavior at Crosswalks

Abstract

For future automated driving functions it is necessary to be able to reason about the typical behavior, intentions and future movements of vulnerable road users in urban traffic scenarios. It is crucial to have this... [ view full abstract ]

Authors

Benjamin Völz (Robert Bosch GmbH, Corporate Research), Holger Mielenz (Robert Bosch GmbH, Corporate Research), Gabriel Agamennoni (ETH Zurich), Roland Siegwart (ETH Zurich)

Topic Areas

Data Mining and Data Analysis , Modeling, Simulation, and Control of Pedestrians and Cyclists , Pedestrian collision avoidance/mitigation

Session

We-B7 » Data Mining and Data Analysis IV (13:40 - Wednesday, 16th September, La Palma)