Papers
5
Total Citations
33
H-Index
4
About
Joonhyung Lee is a robotics researcher advancing human-robot interaction through intelligent, language-guided systems. His primary research areas include interactive robotic agents, semi-autonomous teleoperation, and preference reasoning for object manipulation. Lee’s most impactful contribution is the CLARA framework, which leverages large language models (LLMs) to classify and disambiguate user commands—determining whether they are clear, ambiguous, or infeasible—enabling more reliable and intuitive robot behavior. This work has garnered 24 citations across two versions, highlighting its significance in bridging natural language and robotic control. In the SPOTS project, Lee addresses the underexplored “place” task in pick-and-place operations, developing a teleoperation system that reasons about stable object placement. He also pioneered a quality-diversity approach to semi-autonomous teleoperation, using reinforcement learning to generate diverse robot behaviors that better align with user intentions, overcoming the limitations of homogeneous solution sets. Additionally, his work on visual preference inference allows robots to reason about human preferences from image sequences, enhancing tabletop object manipulation. Through these contributions, Lee is shaping the future of interactive robotics, making autonomous agents more responsive, adaptable, and aligned with human needs.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5