New ANR project on Deep learning for Geophysical Fluid Dynamics

New animations of deep learning application to sea surface dynamics and SWOT data

less than 1 minute read

Published: July 01, 2020

Can we jointly learn variational models and solvers for inverse problems?

less than 1 minute read

Published: June 10, 2020

We've just posted online a new preprint co-authored with Lucas Drumetz and Francois Rousseau on the joint learning of variational models and associated solvers to address inverse problems (e.g., inpainting, denoising, space-time interpolation, data assimilation,...) pdf on Arxiv / Associated code on OceaniX github

New PhD position on Deep learning & Space Oceanography in collaboration with CNES

less than 1 minute read

Published: February 15, 2020

We are pleased to call for applications for a CNES PhD position at IMT Atlantique (Brest, France) on Deep Learning for Space Oceanigraphy newly-funded MELODY project and AI chair OceaniX. More information in the proposed PhD position and the application procedure at: link

Can we get the best of physical models and CNNs for geophysical flows?

less than 1 minute read

Published: February 13, 2020

We've just posted online a new preprint co-authored with Olivier Pannekoucke on PDE-NetGen 1.0: from symbolic PDE representations of physical processes to trainable neural network representations. Arxiv / Associated code