CHUNGBUK NATIONAL UNIV.
INTELLIGENT ROBOTS LAB.

Welcome to the Intelligent Robot Laboratory at Chungbuk National University. We focus on developing next-generation robotic technologies, with core expertise in SLAM, localization, and navigation for autonomous vehicles and AGVs. Our research bridges theory and real-world deployment, enabling intelligent systems to perceive and navigate complex environments. Discover our work through our publications, videos, and ongoing projects.

RECENT NEWS

Featured

We proposed a semantic local and global descriptor-based loop closure detection framework for long-term visual SLAM in challenging urban environments. Our method combines semantically consistent local feature matching with semantically aggregated global scene descriptors in a coarse-to-fine pipeline to achieve robust place recognition under severe illumination, seasonal, viewpoint, and dynamic-object variations. By integrating semantic understanding with handcrafted visual features and fusing local and global similarity scores, the proposed framework improves loop detection accuracy while maintaining real-time performance, outperforming existing state-of-the-art methods across multiple benchmark datasets. Read full paper.

Our latest research on real-time LiDAR loop closure detection for robust place recognition in challenging environments has been published in the Intelligent Service Robotics (ISR) journal. Read here.

Our latest research on a flexible multi-camera visual-inertial odometry system for robust localization in challenging environments has been published in the International Journal of Control, Automation, and Systems (IJCAS) journal. Give it a read here.

A new study from our lab on robust LiDAR-inertial localization for autonomous robots operating in challenging agricultural environments has been published in IEEE Robotics and Automation Letters (RA-L).