Luosong Guo's Homepage
 |
Luosong Guo PhD student East China Normal University(School of Computer Science and Technology) Shanghai Innovation Institute(Collaborative Training) Email: guoluosong123@163.com |
Biography
I am currently a PhD student jointly trained by East China Normal University(ECNU) and Shanghai Innovation Institute(SII), supervised by Prof. Yuan Xie. I received my Master degree from Nanjing University of Aeronautics and Astronautics (NUAA) in 2025, supervised by Prof. Kun Zhu. Prior to that, I completed my Bachelor's degree in the same school at NUAA in 2022. Currently, I am serving as a research intern at Chery, leading a team in conducting research related to autonomous driving closed-loop simulation.
My research interests include Embodied 3D Spatial Intelligence, Autonomous Driving. Specifically, I work on
Perception and Understanding of 3D
Autonomous driving closed-loop simulation
Autonomous driving controllable data generation
News
05/2026 – Congratulation! Our work OmniSceneGraph is now complete. It's an excellent work on reconstruction of a full-scenario autonomous driving environment for physical AI. Looking forward to paper and code release."
01/2026 – Congratulation! Our work MSD3DRES is now complete. It's an excellent work on achieving precise indoor object localization with large 3D models. Looking forward to paper and code release."
01/2026 – Congratulation! Our work HSASCom is accepted by WCNC. It's an excellent work on on semantic communication of 3D point cloud data."
10/2025 – Congratulation! I was invited by the Spatiotemporal Intelligence Department of Chery to give a talk titled "Overview of the Development of 3D Reconstruction Technology for Autonomous Driving". The presentation slides are available here
08/2025 – Congratulation! I joined Chery's Spatiotemporal Intelligence Department as an intern, hoping to achieve exciting achievements here.
07/2025 – Congratulation! We release EmboSceneExplorer. It’s a simulation platform that enables Embodied AI Agents to perform 3D perception and reconstruction, LLM-based grounding, and goal-oriented navigation within virtual 3D scenes. Hava a try now!
07/2025 – Congratulation! I release three tools for indoor semantic occupancy prediction including BlenderCollector, OccGTGenerator, Focus-SAN.
04/2025 – Congratulation! Phd Student life begins, looking forward to more achievements being shared.