Guillaume Cordonnier

Researcher (Chargé de Recherche) at Inria


Office: Y 002 (Byron)
Inria Sophia-Antipolis
2004 route des lucioles, BP 93
FR-06902 Sophia Antipolis, France

Tel: +33 4 92 38 78 40
Mail: guillaume.cordonnier(at)

I am a researcher in the GraphDeco group at Inria. I received my Ph.D. from the Université Grenoble Alpes, supervised by Prof. Marie-Paule Cani and Prof. Eric Galin, where I worked on large scale lansdcape synthesis. Before, I was a student at Ecole Normale Superieur de Lyon, and I received a Master's degree from Grenoble INP and a Bachelor's Degree from the Université de Lyon.

My research interests include simulation and modeling of natural phenomena, and especially how to leverage physically inspired algorithms with intuitive user control.



2021-now Researcher
Inria, GraphDeco
2019-2020 Postdoctoral Researcher
ETH Zurich, Computer Graphics Laboratory
2019 Postdoctoral Researcher
Ecole Polytechnique, STREAM group
2018 Research engineer
Université de Perpignan, CERPT
2015-2018 Ph.D. Student
Université Grenoble Alpes, IMAGINE group


2015-2018 Ph.D. in Computer Science
Université Grenoble Alpes, Inria.
GdR IG-RV Ph.D. award
2015 Master's Degree in Computer Science
Grenoble INP. Specialty Graphics, Vision, Robotics.
2013 Bachelor's Degree in Computer Science
Université de Lyon. Specialty Fundamental Computer Science.
2012-2015 Eleve normalien
Ecole Normale Supérieure de Lyon

  Scientific community

International Program Committee

2019 Motion In Games


2019 Eurographics
Pacific Graphics
Computer & Graphics
2018 Computer Animation and Virtual Worlds

Research projects

Sketch-based design of fluid annimations

Fluids are one of the most common yet challenging effects to model for digital content. The tremendous amount of research in that field brought simulation tools that produce impressive results, but they are unfortunately hard to control. I develop new sketching metaphors for the efficient control of fluid animations while exploring the use of data-driven approaches to reduce the computation time and yield interactive feedback.

Generation of large scale mountain ranges

Current computing capabilities allow for rendering and exploring gigantic digital sceneries. Populating these massive environments increases the need for large scale objects, such as mountain ranges. In this project, I explore the use of geology inspired physical models to generate such content (e.g., fluvial and glacial erosion, plate tectonics and crust compression), and I adapt the resulting algorithms to provide users with interactive and intuitive control.

Interleaved phenomena for landscape synthesis

At medium scales, the visual aspect of landscapes is driven by the competition of interdependent phenomena. I propose new solutions to the efficient combined simulation of numerous complex physical events that operate at various time scales. The applications range from the generation of landscapes including erosion and ecosystems to the simulation of the evolution of snow cover over a season.