Xuhui Kang
Papers
1
Total Citations
7
H-Index
1
About
Xuhui Kang is a rising researcher in robotics and machine learning, whose work centers on enhancing the reliability and robustness of robotic manipulation systems. His most-cited paper, "Diff-Dagger: Uncertainty Estimation With Diffusion Policy for Robotic Manipulation" (2025, 7 citations), tackles a critical challenge in modern robotics: the tendency of diffusion-based policies to fail when encountering out-of-distribution scenarios. Kang’s key contribution lies in integrating uncertainty estimation directly into the diffusion policy framework, enabling robots to recognize when they are operating beyond their training data and to recover from compounding errors. This work bridges a vital gap between state-of-the-art generative modeling and practical, safe deployment in real-world environments. By addressing the fundamental limitations of diffusion policies in extrapolation and failure recovery, Kang’s research has immediate implications for autonomous systems in manufacturing, healthcare, and service robotics. His approach not only improves task success rates but also lays the groundwork for more trustworthy AI-driven manipulation, making him a notable voice in the push toward robust, uncertainty-aware robotic control.
Research Focus
Key Achievements
Top Papers
- 1