Dongjae Lee

I am a first-year Ph.D. student in Mechanical Engineering at the RPM Robotics lab, Seoul National University, advised by Prof. Ayoung Kim.

My research interests include LiDAR-based Simultaneous Localization and Mapping (SLAM), Lifelong Mapping, and Global Localization, with a focus on developing robust and efficient algorithms for autonomous navigation and mapping in dynamic environments. I am particularly interested in leveraging multi-session sensor data with temporal variations for reliable autonomy.

Email  /  Google Scholar  /  Github  /  LinkedIn

ELite is a LiDAR-based lifelong mapping framework that leverages two-stage ephemerality to accurately align multiple sessions, remove dynamic objects, and update maps while robustly distinguishing between transient and persistent environmental changes.

HeLiPR is the first heterogeneous LiDAR dataset designed for place recognition across varying LiDAR types, supporting inter-LiDAR place recognition in diverse environments.

A survey paper on LiDAR odometry, including a comprehensive review of recent advancements and remaining challenges.

Ongoing construction site dataset, supporting the development of robust place recognition algorithms in dynamic, changing environments.

Tightly-coupled GNSS-LiDAR-Inertial state estimator for SLAM and autonomous driving, addressing long-term drift through the integration of raw GNSS measurements, which ensures smooth and accurate state estimation.

Template by Jon Barron.