Workshop on 3D Geometry Generation for Scientific Computing
Answering many of the important scientific questions of our time - "How are ice caps going to melt?"; "How will extreme weather events affect our cities?"; and "How does a forest regrow after fire?" - all require a geometric understanding of the world around us. There have long been efforts to map the world, but current methods in AI & ML allow us to make 3D geometries of outdoor scenes with an ease and fidelity that has never been seen before. With these scenes comes great potential to answer scientific questions of many varieties. This workshop aims to bring together researchers from across AI, computer vision, scientific disciplines, and applied mathematics to discuss the state of the art in 3D geometry generation and the plethora of real-world problems and applications that can use these geometries.Â
We will have speakers across many scientific disciplines including forest ecology, glaciology, and computational fluid dynamics in addition to computer vision experts sharing the state of the art in 3D reconstruction. This workshop aims to provide a forum to share best practices and encourage collaboration among a group of researchers that share a deep interest in reconstructing geometries to answer important scientific questions, but may not have many other opportunities to gather. We especially encourage scientists who have great need for 3D geometries of outdoor scenes, but may not have much experience in computer vision and computer vision researchers in 3D reconstruction who may not have worked in broader science before, but wish to find impactful applications of their work to feel welcome to participate. No prior record of cross-disciplinary work is necessary.
Organizers
Marissa Ramirez de Chanlatte
University of California, Berkeley
marissachanlatte [at] berkeley [dot] edu
Prof. Trevor Darrell
University of California, Berkeley
Prof. Phil Colella
Lawrence Berkeley National Laboratory