Jean Feydy's home page

Preprints

  • Fast end-to-end learning on protein surfaces,
    to be presented at CVPR 2021,
    Freyr Sverrisson*, Jean Feydy*, Bruno Correia, Michael Bronstein, links: Paper, Code.
  • Kernel operations on the GPU, with autodiff, without memory overflows,
    to appear in the Journal of Machine Learning Research,
    Benjamin Charlier*, Jean Feydy*, Joan Glaunès*, François-David Collin, Ghislain Durif, links: Paper, Code.
  • Sinkhorn divergences for unbalanced Optimal Transport,
    submitted in March 2021,
    Thibault Séjourné, Jean Feydy, François-Xavier Vialard, Alain Trouvé, Gabriel Peyré, links: Paper, Code.
  • Collective proposal distributions for nonlinear MCMC samplers: mean-field theory and fast implementation,
    Grégoire Clarté, Antoine Diez, Jean Feydy, links: Paper, Code.

PhD thesis

  • Geometric data analysis, beyond convolutions,
    Jean Feydy, under the supervision of Alain Trouvé.
    Defended on July 2, 2020 in front of Xavier Pennec, Jean-David Benamou, Marc Niethammer, Pierre Alliez and Alexandre Gramfort.
    I was awarded the 2020 PhD thesis award by the AFRIF (French association for shape analysis and recognition).
    Links: Official version, Thesis + Latex, Slides + Latex.

Conference Papers

  • Fast geometric learning with symbolic matrices,
    Jean Feydy*, Joan Glaunès*, Benjamin Charlier*, Michael Bronstein,
    NeurIPS 2020 (spotlight presentation), links: Paper, Slides + Latex, Poster + Latex, Links and Videos, Website, Code.
  • Fast and scalable optimal transport for brain tractograms,
    Jean Feydy*, Pierre Roussillon*, Alain Trouvé, Pietro Gori,
    MICCAI 2019, links: Paper, Poster + PowerPoint, Website, Code.
  • Interpolating between optimal transport and MMD using Sinkhorn divergences,
    Jean Feydy, Thibault Séjourné, François-Xavier Vialard, Shun-ichi Amari, Alain Trouvé, Gabriel Peyré,
    AiStats 2019, links: Paper, Slides, Poster + Latex, Website, Code.
  • Global divergences between measures: from Hausdorff distance to optimal transport,
    Jean Feydy, Alain Trouvé,
    ShapeMI 2018 (Miccai workshop, oral presentation), links: Paper, Slides + Latex, Code.
  • Optimal transport for diffeomorphic registration,
    Jean Feydy, Benjamin Charlier, François-Xavier Vialard, Gabriel Peyré,
    MICCAI 2017 (oral presentation), links: Hal, Arxiv, Code, Slides (+videos), Poster, Latex.
  • Distortion minimizing geodesic subspaces in shape spaces and computational anatomy,
    Benjamin Charlier, Jean Feydy, David W. Jacobs and Alain Trouvé,
    VipImage 2017.

Talks

  • Fast geometric learning with symbolic matrices.
    January 2021, CogSys seminar, DTU Compute (Online): Pdf, Latex.
    December 2020, NeurIPS 2020 (Online): Spotlight presentation + Latex, Poster + Latex, Links and Videos.
  • Geometric data analysis, beyond convolutions.
    April 2021, Signal Processing Laboratory (LTS4, EPFL, Online): Pdf, Latex.
    March 2021, Image, Visual and Language Computing Seminar (UNC Chapel Hill, Online): Pdf, Latex.
    January 2021, Centre de Vision Numérique (CentraleSupélec - Inria Saclay, Online): Pdf, Latex.
    October 2020, Centre de Recherche des Cordeliers (Online) - in French: Pdf, Latex, Video.
    September 2020, University College London (Online) - with more applications: Pdf, Latex, Video, Workshop.
    July 2020, PhD defense (Online): Pdf, Latex.
    December 2019, King's College London: Pdf, Latex.
  • L'imagerie médicale, un calcul structuré.
    October 2020, Institut du Cerveau (Online): Pdf, Latex, Video, Conference.
  • Geometric loss functions for shape analysis
    July 2020, SIAM Imaging Sciences 2020 (Online): Pdf, Video, Latex.
  • Sorting points in dimension D > 1.
    April 2021, Sea Ice Modeling and Data Assimilation (Dartmouth, Online): Pdf, Latex.
    February 2020, Twitter London: Pdf, Latex.
  • Discrete optimal transport: scaling up to 1M samples in 1s.
    June 2019, "People in optimal transportation and applications" workshop, Cortona: Pdf, Latex.
  • Artificial "neural networks": what radiologists should know.
    (French) Mai 2020, DIU Neuro-radiologie vieillissement (en ligne): Pdf, Video;
    (English) June 2019, Harvey Cushing symposium (American Hospital of Paris):
    Pdf 16:9, High-res pptx, Low-res pptx, LaTex source;
    (French) Mars 2019, congrès de la Société Française de Neuro-Radiologie (Paris):
    Pdf 16:9, High-res pptx, Low-res pptx, LaTex source;
    (French) Juin 2018, congrès de la Société Française d'Imagerie Cardiaque et Vasculaire (Beaune):
    Pdf 4:3, Latex + Code.
  • Robust matching of measures with Optimal Transport.
    February 2019, GTTI, ENS Cachan: Pdf, Latex.
    December 2018, BIRS center, Banff: Pdf, Latex.
    November 2018, Télécom ParisTech: Pdf, Latex.
  • Global divergences between measures, from Hausdorff distance to Optimal Transport.
    September 2018, ShapeMI workshop, MICCAI 2018 (Granada): Pdf, Latex + Code.
    July 2018, Curves and Surfaces 2018 (Arcachon): Pdf, Latex + Code.
  • Normalizing LDDMM metrics using autodiff, an introduction to KeOps.
    June 2018, SIAM Imaging Sciences 2018 (Bologna): Pdf, Latex + Code.
    November 2017, Isaac Newton Institute (Cambridge): Pdf, Latex.
  • Automatic differentiation for applied mathematicians: presenting the PyTorch library.
    February 2018, at the ENS Ulm, ENS Cachan and Jussieu (Paris VI): Pdf, Latex + Code.
  • Optimal transport for diffeomorphic registration: a global and robust data attachment term.
    September 2017, MICCAI 2017, Québec City: Pdf (+videos), Poster, Latex.
    June 2017, Asclepios Inria Team: Pdf, Latex.
  • Riemannian geometry for computational anatomy: presenting the LDDMM framework.
    October 2017, ENS Department of Mathematics and Applications: Pdf, Full Latex.
    June 2017, MoKaPlan Inria Team: Pdf (+videos), Latex.