Polynomial Trajectory Predictions For Improved Learning Performance
Ido Freeman, Kun Zhao, Anton Kummert
-
Members: FreeSPS
IEEE Members: $11.00
Non-members: $15.00Length: 00:06:48
20 Sep 2021
The rising demand for Active Safety systems in automotiveapplications stresses the need for a reliable short-term to mid-term trajectory prediction. Anticipating the unfolding path ofroad users, one can act to increase the overall safety. In thiswork, we propose to train neural networks for movement un-derstanding by predicting trajectories in their natural form, asa function of time. Predicting polynomial coefficients allowsus to increase accuracy and improve generalisation.
