Steven Anson Coons Award

The Steven A. Coons award for outstanding creative contributions to computer graphics is presented in odd-numbered years to honor an individual who has made a lifetime contribution to computer graphics and interactive techniques. The award includes a $6,000 cash prize and a specially commissioned statue.

Current Recipient

ACM SIGGRAPH is pleased to present Leonidas J. Guibas with the 2025 Steven Anson Coons Award for Outstanding Creative Contributions to Computer Graphics. He is being recognized for his research contributions to geometry processing and rendering, geometric deep learning, and global illumination — honoring an individual who has made a lifetime contribution to computer graphics and interactive techniques.

Guibas began his research career in theoretical computer science, completing a PhD advised by Donald Knuth at Stanford University in 1976. He initially published important algorithms and data structures for hashing, sorting and searching. These include concepts taught in computer science programs worldwide such as red-black trees for self-balancing binary trees and algorithms for pattern matching in strings.

After his PhD, he worked in industry at the Xerox Palo Alto Research Center (PARC) and then the Systems Research Center, Digital Equipment Corporation, while also joining Stanford University as a Professor of Computer Science. In 1982, Guibas published a paper in the third ever issue of ACM Transactions on Graphics, which started with the line “This paper proposes that bitmaps, or raster images, should be given full citizenship in the world of computer science.” Around this time, Guibas’ attention turned to computational geometry, where his key early results would define the field’s success for decades. Among these were a groundbreaking  divide-and-conquer algorithm for Voronoi diagram construction via its dual Delaunay triangulation, in a paper that also introduced the quad-edge data structure widely used in geometric modeling. In fact, Guibas led an early movement within computational geometry to leverage the power of geometric duality. He also introduced to computational geometry a kinetic framework that, for example, allows determining the swept region of one shape along a curve: a core subroutine of font design and digital drawing. His concept of kinetic data structures, which track attributes of interest in an evolving system over (perhaps virtual) time, has routinely appeared as an ingredient of SIGGRAPH papers for applications in physically based simulation (contact / collision handling) or rendering (visibility queries).

Guibas’s early work on visibility problems within 2D polygons would laid the foundation for computer graphics rendering systems over the next decades. His work on the computational complexity for determining all intersections of a set of curves laid the foundation for powerful geometric algorithm tools such as the CGAL library, a standard tool in geometry processing known for robustness. With his PhD student David Salesin (2000 Computer Graphics Achievement Award), Guibas demonstrated how inexact calculations, using floating point math, could still lead to robust and provably correct geometric constructions. 

Together with his PhD student Eric Veach, Guibas pioneered realistic rendering algorithms that remain central to today’s state of the art. They introduced bidirectional path tracing, which enables tracing rays from both the light source and the eye, addressing the long-standing problem that different optical effects are easier to sample from one or the other. They then introduced multiple importance sampling, which dramatically reduces variance when combining multiple estimators, as is the case with sampling according to a BRDF vs. light source. Finally, they developed Metropolis light transport, which remains the most powerful unbiased approach to the simulation of arbitrary light paths including complex specular effects and caustics — a method extensively in the special effects industry. Veach received a 2014 Academy Scientific and Engineering Award for his PhD work.

In the late 1990s, Guibas made new insights to earth mover’s distance: the problem of moving masses from one distribution to another while minimizing the total distance traveled. This unlocked key metrics for image processing to measure the distance between images or distributions of images, and eventually to the broader community of machine learning as metrics or loss functions for distribution matching. With his PhD student Justin Solomon (2022 Significant New Researcher Award), Guibas demonstrated the power of earth mover’s distance (also referred to as Wasserstein distance or optimal transport) for complex matching problems involving discrete surfaces.

As geometry processing crystallized into its own subcommunity, Guibas repeatedly made key contributions including the heat kernel signature, functional maps and map networks, surface segmentation, and symmetry detection. His functional maps work received the Siggraph 2023 Test-of-Time paper award. Guibas was also at the forefront of machine learning for 3D shapes, making early contributions to convolutional neural networks on 3D shapes and shape retrieval networks. His work on PointNet and PointNet++ was a breakthrough in geometric learning, allowing for the first time to directly train networks over point clouds as input. He initiated and co-developed the ShapeNet dataset, which was the first large scale annotated dataset of 3D models, and stood as the gold standard in the community for over a decade.

Many of Guibas’s trainees are now important faculty and industry leaders in the fields of computer graphics, vision and learning. He has continued to cultivate industrial connections and applications of his work and is currently a principal scientist at Google Deep Mind, in addition to being a Professor at Stanford University. He has been elected a member of the National Academy of Engineering, the National Academy of Sciences, the American Academy of Arts and Sciences, and also as an ACM and IEEE Fellow.

Nomination Procedure

To nominate an individual for the Steven Anson Coons award, ACM SIGGRAPH members are encouraged to send an email to the Technical Awards Chair (technical_awards@siggraph.org) by January 31.

Requirements

Your nomination should describe a candidate’s most significant research contributions, industrial impact, community service, and/or contributions to other areas of computer graphics and interactive techniques. The Technical Awards Committee uses nomination statements as the main basis for their selections, so a concise and clear statement is strongly encouraged. Descriptions of a few most impactful contributions are preferable to a long list of activities.