Kester Duncan
Papers
3
Total Citations
47
H-Index
2
About
Kester Duncan is a researcher whose work sits at the intersection of computer vision, robotics, and human-robot interaction, with a focus on enabling machines to perceive and act upon the physical world. His most impactful contribution is a framework for **multi-scale superquadric fitting**, detailed in his 2013 paper (37 citations), which allows robots to rapidly recover the 3D shape and pose of unknown objects from unorganized point cloud data—a critical capability for autonomous manipulation. This work provides a computationally efficient solution to a core challenge in robotics: handling novel objects without pre-existing models. Duncan has also made significant strides in **scene-dependent intention recognition**, developing systems that allow assistive robots to infer human goals by analyzing environmental context, thereby reducing the need for explicit commands. His 2015 paper on this topic (8 citations) explores how robots can anticipate user actions during collaborative tasks, a key step toward seamless human-robot teamwork. By advancing both geometric perception and social reasoning, Duncan’s research lays essential groundwork for more intuitive, capable, and autonomous robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3