PhD student · Thoth team, Inria Grenoble
I am a PhD student in computer science in the Thoth team at Inria Grenoble, supervised by Julien Mairal and Michael Arbel. My thesis is on foundation models for inverse problems in scientific imaging.
I work on generative priors, in particular diffusion models, for imaging inverse problems, with a focus on Plug-and-Play optimization and the mismatch between MMSE and MAP estimation. More broadly, I am interested in inverse problems, generative modeling and image restoration.
Before starting my PhD, I studied applied mathematics at École Polytechnique and completed the IASD master's in artificial intelligence at PSL University.
Email kenta.vert@inria.fr
Beyond MMSE: Enhancing PnP Restoration with ProxiMAP
Advances in Neural Information Processing Systems (NeurIPS), 2026
@inproceedings{vert2026proximap,
title = {Beyond {MMSE}: Enhancing {PnP} Restoration with {ProxiMAP}},
author = {Vert, Kenta and Meanti, Giacomo and Pesme, Scott and Arbel, Michael and Mairal, Julien},
booktitle = {Advances in Neural Information Processing Systems (NeurIPS)},
year = {2026},
note = {arXiv:2605.16396}
}