## Bio

I am an intern in the computer vision group at ASTRA (previously RITS), a joint team between Inria, Valeo and Valeo.ai. I'm currently working on scene understanding in adverse conditions, under the supervision of Raoul de Charette, Tuan-Hung Vu , Andrei Bursuc and Patrick Pérez. I received an MSc degree in Mathematics, Vision and Learning from ENS Paris-Saclay, an engineering degree from Mines Paris, and a mechanical engineering degree from Lebanese University.

## News

04 / 2022
I started a 6-month internship at Inria under the supervision of Raoul de Charette, Tuan-Hung Vu, Andrei Bursuc and Patrick Pérez.
01 / 2022
LPALM has been accepted at ICLR 2022!
04 / 2021
I started a 6-month internship at the IMAGES team at Télécom Paris under the supervision of Christophe Kervazo, Jérôme Bobin and Florence Tupin.

## Publications

##### Unrolling PALM for Sparse Semi-Blind Source Separation

Mohammad Fahes, Christophe Kervazo, Jérôme Bobin, Florence Tupin
ICLR 2022

@inproceedings{
fahes2022unrolling,
title={Unrolling {PALM} for Sparse Semi-Blind Source Separation},
author={Mohammad Fahes and Christophe Kervazo and J{\'e}r{\^o}me Bobin and Florence Tupin},
booktitle={International Conference on Learning Representations},
year={2022},
url={https://openreview.net/forum?id=aBVxf5NaaRt}
}

Sparse Blind Source Separation (BSS) has become a well established tool for a wide range of applications - for instance, in astrophysics and remote sensing. Classical sparse BSS methods, such as the Proximal Alternating Linearized Minimization (PALM) algorithm, nevertheless often suffer from a difficult hyperparameter choice, which undermines their results. To bypass this pitfall, we propose in this work to build on the thriving field of algorithm unfolding/unrolling. Unrolling PALM enables to leverage the data-driven knowledge stemming from realistic simulations or ground-truth data by learning both PALM hyperparameters and variables. In contrast to most existing unrolled algorithms, which assume a fixed known dictionary during the training and testing phases, this article further emphasizes on the ability to deal with variable mixing matrices (a.k.a. dictionaries). The proposed Learned PALM (LPALM) algorithm thus enables to perform semi-blind source separation, which is key to increase the generalization of the learnt model in real-world applications. We illustrate the relevance of LPALM in astrophysical multispectral imaging: the algorithm not only needs up to $10^{4}-10^{5}$ times fewer iterations than PALM, but also improves the separation quality, while avoiding the cumbersome hyperparameter and initialization choice of PALM. We further show that LPALM outperforms other unrolled source separation methods in the semi-blind setting.