Leonidas Guibas
Papers
1
Total Citations
2
H-Index
1
About
Leonidas Guibas is a pioneering computer scientist whose research spans computational geometry, geometric deep learning, 3D shape analysis, and spatial reasoning. A professor at Stanford University, Guibas has made foundational contributions to how computers perceive, process, and reason about geometric structures — from point clouds and 3D shapes to dynamic physical environments. His work has profoundly influenced fields ranging from robotics and computer vision to graphics and machine learning. Guibas is perhaps best known for advancing the understanding of 3D shape correspondence, part-based shape analysis, and scene understanding. His research group has consistently pushed boundaries in developing algorithms that allow machines to interpret and manipulate the geometry of the physical world. His contributions to data structures in computational geometry, including co-developing the quad-edge data structure, remain foundational references in the field. His recent work, such as "PhysPart: Physically Plausible Part Completion for Interactable Objects" (2024), reflects a continued drive toward practical, real-world applications, bridging generative 3D modeling with physical plausibility for robotics simulation and manufacturing. With an extraordinarily prolific publication record accumulating tens of thousands of citations, Guibas stands as one of the most influential figures in modern geometric computing.
Research Focus
Key Achievements
Top Papers
- 1PhysPart: Physically Plausible Part Completion for Interactable Objects2 citations · 2024