Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Modelling Pedestrians in Autonomous Vehicle Testing

Priisalu, Maria LU (2023) In Doctroral Theses in Mathematical Sciences 2023(4).
Abstract
Realistic modelling of pedestrians in Autonomous Vehicles (AV)s and AV testing is crucial to avoid lethal collisions in deployment. The majority of AV trajectory forecasting literature do not utilize the motion cues present in 3D human pose because it is hard to gather large datasets of articulated 3D pedestrian motion. In this thesis we discuss the difficulties in data gathering and propose a pedestrian model that overcomes the issues by utilizing various datasets and learning paradigms to learn articulated semantically reasoning pedestrian motion. We show that such learnt pedestrian models can and should be utilized in AV testing, instead of heuristics as in previous work, to test AVs on realistic and hard scenarios. We propose a... (More)
Realistic modelling of pedestrians in Autonomous Vehicles (AV)s and AV testing is crucial to avoid lethal collisions in deployment. The majority of AV trajectory forecasting literature do not utilize the motion cues present in 3D human pose because it is hard to gather large datasets of articulated 3D pedestrian motion. In this thesis we discuss the difficulties in data gathering and propose a pedestrian model that overcomes the issues by utilizing various datasets and learning paradigms to learn articulated semantically reasoning pedestrian motion. We show that such learnt pedestrian models can and should be utilized in AV testing, instead of heuristics as in previous work, to test AVs on realistic and hard scenarios. We propose a framework for generating varied AV test scenarios by posing AV test case generation as a visual problem. Finally we provide a method to improve existing articulated human pose forecasting by utilizing individual specific motion cues on the fly. This thesis discusses the difficulties in articulated pedestrian sensing, proposes a pedestrian model to overcome these difficulties showing a direct use of the pedestrian model in AV testing, and shows the possible further improvements to articulated pedestrian motion forecasting should articulated models be utilized in AV trajectory planning. We hope that this work aids in the further development of articulated and semantically reasoning pedestrian models in AV testing and trajectory planning. (Less)
Please use this url to cite or link to this publication:
author
supervisor
opponent
  • Doc. Ahlberg, Jörgen, Linköping University, Sweden.
organization
publishing date
type
Thesis
publication status
published
subject
keywords
pedestrian sensing, pedestrian forecasting, pedestrian motion synthesis, generative testing, autonomous vehicle testing, reinforcement learning
in
Doctroral Theses in Mathematical Sciences
volume
2023
issue
4
pages
232 pages
publisher
Centre for Mathematical Sciences, Lund University
defense location
Lecture Hall MA3, Centre of Mathematical Sciences, Sölvegatan 20, Faculty of Engineering LTH, Lund University, Lund. The dissertation will be live streamed, but part of the premises is to be excluded from the live stream.
defense date
2023-11-06 13:00:00
ISSN
1404-0034
ISBN
978-91-8039-827-5
978-91-8039-828-2
project
Modelling Pedestrians in Autonomous Vehicle Testing
Semantic Mapping and Visual Navigation for Smart Robots
language
English
LU publication?
yes
id
a65bb853-a440-4418-b92a-f74049a9d9dd
date added to LUP
2023-10-09 14:36:28
date last changed
2023-10-12 09:33:07
@phdthesis{a65bb853-a440-4418-b92a-f74049a9d9dd,
  abstract     = {{Realistic modelling of pedestrians in Autonomous Vehicles (AV)s and AV testing is crucial to avoid lethal collisions in deployment. The majority of AV trajectory forecasting literature do not utilize the motion cues present in 3D human pose because it is hard to gather large datasets of articulated 3D pedestrian motion. In this thesis we discuss the difficulties in data gathering and propose a pedestrian model that overcomes the issues by utilizing various datasets and learning paradigms to learn articulated semantically reasoning pedestrian motion. We show that such learnt pedestrian models can and should be utilized in AV testing, instead of heuristics as in previous work, to test AVs on realistic and hard scenarios. We propose a framework for generating varied AV test scenarios by posing AV test case generation as a visual problem. Finally we provide a method to improve existing articulated human pose forecasting by utilizing individual specific motion cues on the fly. This thesis discusses the difficulties in articulated pedestrian sensing, proposes a pedestrian model to overcome these difficulties showing a direct use of the pedestrian model in AV testing, and shows the possible further improvements to articulated pedestrian motion forecasting should articulated models be utilized in AV trajectory planning. We hope that this work aids in the further development of articulated and semantically reasoning pedestrian models in AV testing and trajectory planning.}},
  author       = {{Priisalu, Maria}},
  isbn         = {{978-91-8039-827-5}},
  issn         = {{1404-0034}},
  keywords     = {{pedestrian sensing; pedestrian forecasting; pedestrian motion synthesis; generative testing; autonomous vehicle testing; reinforcement learning}},
  language     = {{eng}},
  month        = {{10}},
  number       = {{4}},
  publisher    = {{Centre for Mathematical Sciences, Lund University}},
  school       = {{Lund University}},
  series       = {{Doctroral Theses in Mathematical Sciences}},
  title        = {{Modelling Pedestrians in Autonomous Vehicle Testing}},
  url          = {{https://lup.lub.lu.se/search/files/161296804/Maria_Thesis_electronic.pdf}},
  volume       = {{2023}},
  year         = {{2023}},
}