Yonghyeon Lee

Seoul National University, Korea Institute for Advanced Study

Papers

Total Citations

H-Index

About

Yonghyeon Lee is a robotics researcher whose work bridges the gap between perception and manipulation, with a focus on enabling robots to interact with the physical world in more adaptive and intelligent ways. His key research areas include robotic grasping, motion planning, and representation learning for manipulation tasks. Lee's most notable contribution is the development of DSQNet (2022, 19 citations), a deformable model-based supervised learning algorithm that tackles the challenging problem of grasping unknown, partially occluded objects. This work is particularly significant because it addresses a critical limitation of deep learning-based end-to-end methods, which often require extensive training data and computational resources. By proposing a more practical and efficient approach, Lee's research has implications for real-world robotic applications where data and resources are limited. In his more recent work, MMP++ (2024, 5 citations), Lee advances the concept of motion manifold primitives by incorporating parametric curve models, enhancing the system's ability to generate diverse and adaptable trajectories. This innovation addresses a key limitation of previous models, which lacked crucial temporal and functional capabilities. Through these contributions, Lee is helping to create more versatile and robust robotic systems capable of operating in unstructured environments.