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

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Mahjourian, Reza, Mu, Rongbing, Likhosherstov, Valerii, Mougin, Paul, Huang, Xiukun, Messias, Joao, Whiteson, Shimon
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929336819384320
author Mahjourian, Reza
Mu, Rongbing
Likhosherstov, Valerii
Mougin, Paul
Huang, Xiukun
Messias, Joao
Whiteson, Shimon
author_facet Mahjourian, Reza
Mu, Rongbing
Likhosherstov, Valerii
Mougin, Paul
Huang, Xiukun
Messias, Joao
Whiteson, Shimon
contents 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 existing scene. Our unified modeling approach, combined with autoregressive agent injection, conditions the placement and motion trajectory of every new agent on all existing agents and their trajectories, leading to realistic scenarios with low collision rates. Our experimental results show that UniGen outperforms prior state of the art on the Waymo Open Motion Dataset.
format Preprint
id arxiv_https___arxiv_org_abs_2405_03807
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle UniGen: Unified Modeling of Initial Agent States and Trajectories for Generating Autonomous Driving Scenarios
Mahjourian, Reza
Mu, Rongbing
Likhosherstov, Valerii
Mougin, Paul
Huang, Xiukun
Messias, Joao
Whiteson, Shimon
Robotics
Machine Learning
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 existing scene. Our unified modeling approach, combined with autoregressive agent injection, conditions the placement and motion trajectory of every new agent on all existing agents and their trajectories, leading to realistic scenarios with low collision rates. Our experimental results show that UniGen outperforms prior state of the art on the Waymo Open Motion Dataset.
title UniGen: Unified Modeling of Initial Agent States and Trajectories for Generating Autonomous Driving Scenarios
topic Robotics
Machine Learning
url https://arxiv.org/abs/2405.03807