Seung-Hun Jeon
Papers
5
Total Citations
108
H-Index
4
About
Seung-Hun Jeon is a leading roboticist whose work bridges the frontiers of legged locomotion, whole-body manipulation, and creative machine intelligence. His core research focuses on enabling robots to perceive and interact with the physical world with unprecedented autonomy. Jeon’s most impactful contribution is a state estimation algorithm for legged robots, which reformulates the problem as a Maximum A Posteriori (MAP) estimation solved via the Gauss-Newton method, incorporating Schur complement marginalization for computational efficiency (45 citations). He further advanced the field by developing a learning-based hierarchical control system that allows quadrupedal robots to manipulate large, heavy objects using their entire body, a breakthrough that captures manipulation-relevant information through deep latent variable embeddings (40 citations). Jeon’s work extends beyond traditional robotics into creative domains, where he has pioneered robot learning from demonstrations for painting tasks, tackling the stochastic dynamics of physical contact and color blending. His recent innovations include a proprioceptive state estimator combining invariant extended Kalman filters with neural measurement networks, ensuring reliable robot operation in vision-denied environments. With over 100 total citations and a trajectory that spans from foundational state estimation to artistic expression, Jeon is shaping the next generation of perceptive, dexterous, and creative robotic systems.
Research Focus
Key Achievements
Top Papers
- 1Legged Robot State Estimation With Dynamic Contact Event Information45 citations · 2021
- 2Learning Whole-Body Manipulation for Quadrupedal Robot40 citations · 2023
- 3
- 4Robot Learning to Paint from Demonstrations4 citations · 2022
- 5