Saved in:
Bibliographic Details
Main Authors: Ding, Wenhao, Cao, Yulong, Zhao, Ding, Xiao, Chaowei, Pavone, Marco
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2312.13303
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909285508710400
author Ding, Wenhao
Cao, Yulong
Zhao, Ding
Xiao, Chaowei
Pavone, Marco
author_facet Ding, Wenhao
Cao, Yulong
Zhao, Ding
Xiao, Chaowei
Pavone, Marco
contents Simulation plays a crucial role in the development of autonomous vehicles (AVs) due to the potential risks associated with real-world testing. Although significant progress has been made in the visual aspects of simulators, generating complex behavior among agents remains a formidable challenge. It is not only imperative to ensure realism in the scenarios generated but also essential to incorporate preferences and conditions to facilitate controllable generation for AV training and evaluation. Traditional methods, mainly relying on memorizing the distribution of training datasets, often fall short in generating unseen scenarios. Inspired by the success of retrieval augmented generation in large language models, we present RealGen, a novel retrieval-based in-context learning framework for traffic scenario generation. RealGen synthesizes new scenarios by combining behaviors from multiple retrieved examples in a gradient-free way, which may originate from templates or tagged scenarios. This in-context learning framework endows versatile generative capabilities, including the ability to edit scenarios, compose various behaviors, and produce critical scenarios. Evaluations show that RealGen offers considerable flexibility and controllability, marking a new direction in the field of controllable traffic scenario generation. Check our project website for more information: https://realgen.github.io.
format Preprint
id arxiv_https___arxiv_org_abs_2312_13303
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle RealGen: Retrieval Augmented Generation for Controllable Traffic Scenarios
Ding, Wenhao
Cao, Yulong
Zhao, Ding
Xiao, Chaowei
Pavone, Marco
Machine Learning
Artificial Intelligence
Simulation plays a crucial role in the development of autonomous vehicles (AVs) due to the potential risks associated with real-world testing. Although significant progress has been made in the visual aspects of simulators, generating complex behavior among agents remains a formidable challenge. It is not only imperative to ensure realism in the scenarios generated but also essential to incorporate preferences and conditions to facilitate controllable generation for AV training and evaluation. Traditional methods, mainly relying on memorizing the distribution of training datasets, often fall short in generating unseen scenarios. Inspired by the success of retrieval augmented generation in large language models, we present RealGen, a novel retrieval-based in-context learning framework for traffic scenario generation. RealGen synthesizes new scenarios by combining behaviors from multiple retrieved examples in a gradient-free way, which may originate from templates or tagged scenarios. This in-context learning framework endows versatile generative capabilities, including the ability to edit scenarios, compose various behaviors, and produce critical scenarios. Evaluations show that RealGen offers considerable flexibility and controllability, marking a new direction in the field of controllable traffic scenario generation. Check our project website for more information: https://realgen.github.io.
title RealGen: Retrieval Augmented Generation for Controllable Traffic Scenarios
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2312.13303