GDR IASIS workshop on flow matching and diffusion
Due to the GDR IASIS website being down until Sept 17th, this is a temporary page to host the workshop’s info.
Registration (in-person or remote) is free but mandatory here: https://forms.gle/Sz4F2sXBjzFzs2wo9
Call for submissions
The call for submissions (20 min oral or poster) is open, send your title+abstract proposals to both mathurin.massias@gmail.com and segolene.martin@inria.fr
Practical info
- Location: Amphi D2, ENS de Lyon (Descartes site, 15 parvis René Descartes, 69007 Lyon - Debourg metro station)
- Date: October 9th 2026, from 09:00 to 17:30. lunch break on you own 12 h 15 - 13 h 45, poster session 13 h 45 - 15 h 15
- For remote participants, a zoom link will be made available a few days before the workshop
- Organizers: Ségolène Martin and Mathurin Massias
Scientific content
Generative models have seen spectacular recent advances, to the point where their latest versions are now capable of producing synthetic images and text that are nearly indistinguishable from real content. Among the approaches that have contributed to this progress, diffusion and flow matching methods occupy a central place.
This meeting aims to take stock of recent advances, both fundamental and applied, in the field of generative models, with a particular emphasis on applications in image processing. The call for contributions is open to work on theoretical developments, algorithmic developments, or original applications. A non-exhaustive list of topics includes:
- generalization of diffusion and flow matching methods
- training of generative models
- scaling up generation
- diffusion and flow matching for discrete data and text
- connections with optimal transport
- evaluation of generative models
- integration of generative models into image processing applications
Invited speaker (50 min talks):
- Anna Korba – TBA
- Vicky Kalogeiton (LIX, Ecole Polytechnique) – Scale is religion?
- Umut Simsekli (SIERRA, INRIA) – TBA
This event benefits from the support of the Institut Rhonalpin des Systèmes complexes (IXXI)