Max Robotics — Global Service Robot Directory + US Certification

About

Suyong Yeon is a robotics researcher specializing in 3D perception, place recognition, and simultaneous localization and mapping (SLAM) for autonomous navigation in complex indoor environments. His most impactful work, "SpoxelNet" (2020, 29 citations), introduced a novel spherical voxel-based deep learning approach for place recognition in crowded indoor spaces using 3D point clouds, addressing a critical challenge for full robot autonomy. Yeon has made foundational contributions to observability analysis in SLAM, developing condition number-based methods to quantify how much of a system is observable—a key advancement for state estimation in noisy real-world environments. His research spans object-oriented 3D RGB-D mapping for realistic indoor reconstruction and image-based localization using wayfinding maps, bridging the gap between human-readable abstractions and robot navigation. Yeon has also contributed large-scale localization datasets specifically designed for crowded indoor spaces, enabling benchmarking and advancement in the field. His most recent work, "Mode-GS" (2024), tackles novel-view rendering for ground-robot trajectories using monocular depth-guided anchored 3D Gaussian splatting, overcoming limitations of prior neural rendering methods. Through these contributions, Yeon continues to advance the reliability and robustness of autonomous navigation systems in challenging indoor environments.