The Mi³ Lab is co-organizing the DriveX Workshop on Foundation Models for V2X-Based Cooperative Autonomous Driving in conjunction with CVPR 2025, taking place June 11–15, 2025 in Nashville, USA.
DriveX will bring together researchers and practitioners to explore how emerging foundation models and vehicle-to-everything (V2X) systems can improve autonomous vehicle perception, planning, and decision-making. The workshop addresses challenges such as holistic scene understanding, 3D occupancy prediction, cooperative perception, and safety-driven multi-agent collaboration.
Topics of Interest Include:
- Foundation Models for Cooperative Autonomous Driving
- Vision-Language Models (VLMs) for Traffic Scene Understanding
- Large Language Model (LLM)-Assisted Cooperative Systems
- Communication-Efficient Cooperative Perception
- Dataset Curation and Benchmarking for AV Systems
- 3D Object Detection, Semantic Segmentation, and Occupancy Prediction
- End-to-End Systems and V2I Interaction
- Safety and Standards in Autonomous Driving
Workshop Highlights:
- Organizers include Mi³ Lab Director Ross Greer (UC Merced), along with Walter Zimmer (Technical University of Munich), Max Ronecker (Graz University of Technology), Chuheng Wei (UC Riverside), and a diverse team from UC San Diego, UCLA, Univerity of Hong Kong, University of Sydney, and more. See full organizing committee.
- Keynote Speakers include Prof. Trevor Darrell (UC Berkeley), Prof. Wolfram Burgard (UTN), Prof. Daniel Cremers (TUM), Dr. Mingxing Tan (Waymo), Prof. Marco Pavone (Stanford/NVIDIA), Prof. Laura Leal-Taixé (TUM/NVIDIA), and others. See full list.
- The DriveX Challenge is a 10-week competition based on the TUMTraf V2X Dataset, focused on cooperative 3D object detection and tracking. Top teams will be invited to present and receive recognition.
-
Awards will be given for Best Paper, Best Student Paper, Best Application Paper, and Challenge Winners — with cash prizes and formal certificates.
Learn more at the DriveX Workshop website.