Xuan Han

Papers

Total Citations

H-Index

About

Xuan Han is a leading researcher in robotic assembly and 3D geometric understanding, with a focus on enabling autonomous systems to reason about and manipulate object structures. Their most cited work, "3D Part Assembly Generation With Instance Encoded Transformer" (2022, 19 citations), tackles the challenging problem of furniture assembly from complete part geometries—a task that demands deep structural comprehension of how components fit together in 3D space. By introducing an instance-encoded transformer architecture, Han’s work moves beyond simple part recognition to model the spatial relationships and assembly sequences required for robots to perform complex construction tasks autonomously. This contribution is foundational for advancing robotic manipulation in manufacturing, home assistance, and disaster response, where assembling objects from parts is critical. Han’s research bridges computer vision, robotics, and generative modeling, offering a novel framework that learns to predict 6-DoF poses for each part, ensuring coherent assembly. With growing interest in embodied AI and automated manufacturing, Han’s work is poised to influence future systems that can understand and build physical structures from digital models.