Homepage

💼 Short Bio

My name is Nikolaos Stathoulopoulos (Nikos for short) and I am a Robotics & AI Ph.D. candidate at Luleå University of Technology in Sweden. I am working at the intersection of localization, mapping, and 3D perception. My research focuses on multi-robot/multi-session SLAM (Simultaneous Localization and Mapping) and long-term map management, often combining classical techniques with learning-based methods to make perception stacks more adaptive and reliable in the field. I care about systems that last, maps that age gracefully, scale efficiently, remain robust under change, and ultimately stand the test of time. Core topics include place recognition, loop closures, and back-end optimization for robust, lifelong operation.

I hold a Bachelor's with an integrated Master in Electrical and Computer Engineering from the University of Patras in Greece. During my current position as a Ph.D. student I also serve as a Teaching Assistant for Advanced Robotics and Computer Vision & Image Processing. I’m active in the IEEE Robotics & Automation Society (RAS) and the broader robotics community, with publications in ICRA, IROS, RA-L and T-FR. I’m always open to collaboration, code sharing, and brainstorming, feel free to reach me.

🗞️ Latest News

— 2026 —

[2026-01-02] The code and pretrained weights for the Scene Graph-Aware Deep Point Cloud Compression (SGA‑DPCC) are now available on the project’s GitHub: LTU-RAI/sga-dpcc.

— 2025 —

[2025-11-24] Our article on keyframe sampling for large-scale SLAM has been accepted for publication in the IEEE Robotics and Automation Letters (RA-L) and will also be presented in the upcoming 2026 IEEE International Conference on Robotics and Automation (ICRA) in Vienna!! Code will be realesed by February 2026.
IEEExplore | arXiv | YouTube

[2025-11-20] I recently gave three lectures on SLAM for the R7021E Advanced Robotics course at LTU. The material includes detailed Jupyter notebooks and Python/pygame examples to make SLAM concepts accessible and hands-on. Check out the project here: hello-slam.

[2025-10-17] Our article on scene graph-aware point cloud compression has been accepted for publication in the IEEE Robotics and Automation Letters (RA-L) and will also be presented in the upcoming 2026 IEEE International Conference on Robotics and Automation (ICRA) in Vienna!! Code will be realesed by January 2026.
IEEExplore | arXiv | YouTube

— 2024 —

[2024-11-20] Our project has been listed in IVA’s 100 List of the Royal Swedish Academy of Engineering Sciences, as part of the current research with potential to create value!!
IVA List | Project

[2024-10-14] We won the Best Paper Award in the “Standing the Test of Time: Retrospective and Future of World Representations for Lifelong Robotics” workshop at IROS’24!!
Workshop | Paper | LinkedIn

[2024-06-28] Our journal extension of FRAME has been accepted for publication in the IEEE Transactions of Field Robotics (T-FR). Luckily, it will be the first page of the first ever volume!!
IEEExplore | arXiv