Driving into the Future: 3D Vision Perception for Autonomous Driving
CVML 2026 Workshop – September 18, 2026 (Nanchang, China)

About the Workshop
We are thrilled to invite you to the CVML 2026 Workshop on 3D Vision Perception for Autonomous Driving – a focused, high‑energy satellite event taking place on September 18, 2026 (the first day of the main conference) in Nanchang, China.
This workshop zeroes in on the most challenging and exciting frontiers in autonomous driving perception: from LiDAR point clouds and multi‑view stereo to dynamic scene understanding and sensor fusion. Our goal is to bridge cutting‑edge research with real‑world deployment, offering a vibrant stage for researchers, engineers, and industry pioneers to exchange breakthrough ideas, tackle the complexities of urban traffic, and push the boundaries of safe, high‑level autonomy.

Topics of Interest (including but not limited to)

⭐Deep learning‑based point cloud segmentation, detection, and tracking
⭐Monocular/binocular depth estimation & dense 3D reconstruction
⭐Multi‑sensor (camera, LiDAR, radar) fusion for robust perception
⭐ Dynamic object behavior prediction and interaction modeling
⭐ Lightweight 3D perception models and edge‑computing solutions
⭐ Domain adaptation from simulation to real‑world scenarios
⭐ Autonomous driving datasets, benchmarking, and evaluation
⭐ Visual representation learning for end‑to‑end autonomous driving

📅 Important Dates (All times are midnight AoE)
🗓️Full paper submission deadline | July 14, 2026 |
🗓️Notification of acceptance | August 28, 2026 |
🗓️Author registration deadline | September 11, 2026 |
🗓️Workshop date | September 18, 2026|

📄 Paper Submission
All submissions must be made via the Microsoft CMT system (shared with the main conference). Please select the track “Workshop on 3D Vision Perception”when submitting.
Submission link: [https://cmt3.research.microsoft.com/CVML2026](https://cmt3.research.microsoft.com/CVML2026)
) ✅ Accepted and registered papers will be published alongside main‑conference papers in the CVML 2026 proceedings by SPIE (ISSN: 0277‑786X) and submitted to EI Compendex & Scopus for indexing.

Organization & Support
🏛️ Organized by Jiangxi Normal University**, China
🤝 Supported by Xi’an Jiaotong University & Jiangxi University of Science and Technology