Myung–Hwan Jeon

Institute for Advanced Engineering, Korea Advanced Institute of Science and Technology

Papers

Total Citations

H-Index

About

Myung-Hwan Jeon is a leading researcher in robotic perception and computer vision, with a focus on transparent object recognition and 6D pose estimation. His work addresses fundamental challenges in visually-assisted robot manipulation, particularly where conventional RGB and depth sensors fail. Jeon’s key contributions include the development of the TRansPose dataset, a large-scale multispectral benchmark that leverages thermal infrared imaging to detect and recognize transparent objects—a notoriously difficult problem due to their material properties. This dataset has become a critical resource, garnering 11 citations since its 2023 release. In parallel, Jeon advanced ambiguity-aware multi-object pose optimization, introducing uncertainty-aware frameworks that handle occlusion and structural symmetry in 6D pose estimation. His 2022 paper on this topic, with 9 citations, directly addresses the limitations of prior work that demanded deterministic predictions in ambiguous scenarios. By enabling robots to reason about pose uncertainty, Jeon’s research enhances the reliability of autonomous manipulation in cluttered, real-world environments. His work bridges a crucial gap between perception and action, making him a notable figure in the intersection of robotics and computer vision.