Currently the maps in PGDrive are obtained by Procedural Generation, and the cruise behavior of traffic vehicles is powered by the
IDM (Intelligent Driving model). Can we take advantage of real-world datasets, such as Argoverse, by offering the transformation for maps and vehicle behaviors from the real world to the simulator?
Currently the maps in PGDrive are obtained by Procedural Generation, and the cruise behavior of traffic vehicles is powered by the IDM (Intelligent Driving model). Can we take advantage of real-world datasets, such as Argoverse, by offering the transformation for maps and vehicle behaviors from the real world to the simulator?