Schedule

The workshop will take place at ENS Lyon (France), Sep 14–15 (2026).

Day & Date Time Talk (Speaker: TBC)
Monday
14 September
09:30 - 10:00 Welcome and coffee
10:00 - 10:50 Mike Davies — Equivariant Imaging: Theory, Algorithms and Applications
10:50 - 11:40 Florence Forbes — Scalable Bayesian Experimental Design with Diffusions
11:45 - 14:00 Lunch + Poster Session
14:00 - 15:00 Rémi Gribonval — Training dynamics of ReLU Networks: a Path-lifting Perspective
15:00 - 16:00 Caroline Chaux — Learning through unrolling a weighted least squares data term to adapt to noise in image deconvolution
16:00 - 16:30 Coffee break
16:30 - 17:30 Julien Mairal — Machine learning and optimization for scientific imaging
Tuesday
15 September
09:30 - 10:30 Rich Baraniuk — Learning from Collapse: Self-Consuming Loops in AI and the Humans Who Feed Them
10:30 - 10:50 Coffee break
10:50 - 11:40 Marcelo Pereyra — Physics-Grounded Foundation Models for Generative Computational Imaging
11:45 - 14:30 Lunch and Poster Session
14:30 - 16:30 Julian Tachella — Self-supervised learning from noisy and incomplete data (HdR presentation)