[Paper] MPC-based Imitation Learning for Human-like Autonomous Driving

We have been doing research on learning from human driving to improve autonomous driving comfort and safety for a while. This paper draft is a nice piece of writing with quite some works, discussions, demonstrations, and evaluations on different learning methods. In particular, we also show an efficient hierarchical learning architecture to combine imitation learning (IL) and differentiable model predictive control (MPC). If you are interested in ADAS or AV, this is an interesting reading.

Link: https://lnkd.in/euExfmct

with Flavia Sofia Acerbo (industrial PhD at Siemens Digital Industries Software), Prof. Tinne Tuytelaars (ESAT, KU Leuven), and Prof. Jan Swevers (Mech Dept., KU Leuven).