Mattéo Clémot

PhD student in computer science

I am a PhD student in topological data analysis at the Université Lyon 1.

Education

2023–2026 Ph.D. student in computer science
Dimensionality reduction under topological constraints
Supervised by Julie Digne (LIRIS) & Julien Tierny (LIP6)
Part of the project TORI
2019–2022 Bachelor's and master's degree in fundamental computer science
ENS de Lyon
2016–2018 Classe préparatoire aux grandes écoles, MPSI - MP*
Lycée Condorcet, Paris

Publications

M. Clémot, J. Digne, J. Tierny, Topological Autoencoders++: Fast and Accurate Cycle-Aware Dimensionality Reduction. Preprint.
arXiv Code
D. Attali, M. Clémot, B. B. Dornelas, and A. Lieutier, When alpha-complexes collapse onto codimension-1 submanifolds. Accepted at SOCG 2025.
arXiv
M. Clémot and J. Digne, Neural skeleton: implicit neural representation away from the surface, Computer & Graphics, 2023. Presented at SMI 2023.
HAL Code GRSI Slides

Research internships

Feb.–Apr.
2023
Vector-field design on implicit SDFs
Supervised by Amir Vaxman
Institute of Perception, Action and Behaviour, The University of Edinburgh
Oct.–Dec.
2022
Solving ill-posed inverse problems involving partial differential equations with neural networks
Supervised by Elisa Riccietti & Stefania Bellavia
Team DANTE, LIP, Lyon
Report Slides
Report
Slides
Feb.–July
2022
Learning the topology of shapes given by point clouds
Supervised by Julie Digne
Team Origami, LIRIS, Lyon
Report Slides Code
Report
Slides
Code
Apr.–July
2021
Surface reconstruction with Delaunay complexes
Supervised by Dominique Attali
Team GAIA, GIPSA-lab, Grenoble
Report Slides Code
Report
Slides
Code
June–July
2020
Links between tropical linear programming, mean payoff games and parity games
Supervised by Stéphane Gaubert & Xavier Allamigeon
Team Tropical, INRIA Saclay & CMAP (École polytechnique), Palaiseau
Report Slides
Report
Slides