UniGen: Unified Modeling of Initial Agent States and Trajectories for Generating Autonomous Driving Scenarios

arXiv:2405.03807v1 Announce Type: cross
Abstract: This paper introduces UniGen, a novel approach to generating new traffic scenarios for evaluating and improving autonomous driving software through simulation. Our approach models all driving scenario elements in a unified model: the position of new agents, their initial state, and their future motion trajectories. By predicting the distributions of all these variables from a shared global scenario embedding, we ensure that the final generated scenario is fully conditioned on all available context in the …

abstract agent agents arxiv autonomous autonomous driving cs.lg cs.ro driving improving modeling novel paper simulation software state through traffic type unified model