Sungwook Lee
Papers
1
Total Citations
7
H-Index
1
About
Sungwook Lee is a researcher at the forefront of robotic manipulation, with a primary focus on developing robust and reliable learning-based control systems. His key research areas include imitation learning, uncertainty estimation, and diffusion models for robotics. Lee’s most notable contribution is his work on "Diff-Dagger," a novel framework that integrates uncertainty estimation with diffusion policy to address critical limitations in robotic manipulation. While diffusion policies have shown promise in handling multi-modal tasks, they suffer from compounding errors and poor out-of-distribution generalization. Lee’s approach directly tackles these challenges by enabling robots to recognize when they are operating outside their training distribution, thereby reducing catastrophic failures. With 7 citations since its 2025 publication, this work is already recognized as a significant step toward safer, more dependable autonomous systems. Lee’s research is particularly impactful for students and engineers working on real-world deployment of robotic systems, where reliability is paramount. His contributions represent a meaningful advance in bridging the gap between powerful generative models and practical, uncertainty-aware robotic control.
Research Focus
Key Achievements
Top Papers
- 1