This invited session aims to bring together leading researchers and engineers from academia and industry to discuss the latest advances in autonomous driving. The session is now in its third year (see our event last year) and has been continuously evolving with this rapidly changing field, covering all areas of autonomy including perception, behavior prediction, and motion planning.
Machine learning-based methods show promise in addressing the fundamental challenges of autonomous driving and can significantly enhance the performance, scalability, and intelligence of the system, allowing autonomous vehicles to operate in complex driving environments. The use of machine learning-based methods for perception, prediction, and decision-making in autonomous driving is a rapidly growing area of research, with a scope that extends from BEV perception, occupancy prediction, online map building, behavior prediction, and motion planning.
This year, the focus is specifically on building and utilizing large foundation models for autonomous driving-related tasks, including scene understanding, prediction, and decision-making. Large, generally trained models contain a wealth of information, and harnessing such a breadth of knowledge in autonomous driving can help autonomous vehicles generalize to new domains and navigate smoothly in rare scenarios. In addition to this, the session is also interested in novel end-to-end driving architectures, traffic simulation, AI safety, and human-machine interaction.
We are soliciting original contributions that are not published or currently under consideration by any other journals/conferences.
Foundational models for autonomous driving.
Vision language models (VLMs) and large language models (LLMs) for solving autonomous vehicle related tasks such as perception, prediction, or planning.
End-to-end autonomous driving architectures.
Bird’s eye view methods for autonomous driving, such as BEV-based 3D detection, BEV segmentation, occupancy grids, HD-maps, and topological lane graphs.
Prediction for model-based decision-making, planning, and control.
Behavior/intention prediction for heterogeneous traffic participants.
Simulation for autonomous driving.
Reinforcement learning, imitation learning, and inverse reinforcement learning.
Diffusion models for prediction, planning, and simulation.
Human-in-the-loop learning and learning from human feedback.
Human-machine interaction and human-AI collaboration.
Verification and validation of learning-based systems.
System safety and cyber security of autonomous systems.
The expected length of the manuscript is 6 pages. A page charge needs to be paid if your final paper is over this page limit. A maximum of 2 additional pages is allowed, but at an extra cost per page. The maximum number of pages is 6 + 2 (with additional cost) = 8. Each paper will undergo a peer-reviewing process by at least two independent reviewers. All the accepted papers, if they are presented at ITSC 2024, will be published in IEEE Xplore and eligible for the journal special issues arranged for the conference.
The submission portal (https://its.papercept.net/conferences/scripts/start.pl) is now open and authors can submit special session papers through the link.
Select Submit a contribution to ITSC 2024.
Submit as an Invited Session Paper with code dfg21.
May 01, 2024: Submission deadline for invited session papers
July 07, 2024: Notification of paper acceptance
July 31, 2024: Final paper submission deadline
TBD
Nanyang Technological University
Nanyang Technological University
Nanyang Technological University
University of California, Berkeley
KTH Royal Institute of Technology, Sweden
Nanyang Technological University