I am a Research Scientist at Owkin, working on computer vision for histopathology. My research spans self-supervised learning, spatial transcriptomics and cutting-edge imaging modalities like imaging mass cytometry (IMC) and immunofluorescence (IF).

I am a core contributor to skglm, a solver for high-dimensional sparse learning problems and, ggml, a C library for efficient tensor calculus. I have created bark.cpp, a package to run text-to-speech models on the edge.

Previously, I was an ML researcher at French computer science research institute INRIA in the Parietal team, working on sparse neuroimaging models and optimization with Alexandre Gramfort and Joseph Salmon.

For my master thesis, I worked with Mathurin Massias on convex and non-convex non-smooth optimization algorithms.

I hold an MSc in Data Science from Ecole Polytechnique. Before that, I graduated from HEC Paris (management) in France.

publications

  1. AI-based identification of FGFR3 mutation status from routine histology slides of muscle-invasive bladder cancer.
    Charlie Saillard, Pierre-Antoine Bannier, Philipp Mann, and 4 more authors
    Journal of Clinical Oncology, 2023
  2. Beyond L1: Faster and Better Sparse Models with skglm
    Quentin Bertrand, Quentin Klopfenstein, Pierre-Antoine Bannier, and 2 more authors
    Advances in Neural Information Processing Systems (NeurIPS), 2022
  3. Benchopt: Reproducible, efficient and collaborative optimization benchmarks
    Thomas Moreau, Mathurin Massias, Alexandre Gramfort, and 18 more authors
    Advances in Neural Information Processing Systems (NeurIPS), 2022
  4. Electromagnetic neural source imaging under sparsity constraints with SURE-based hyperparameter tuning
    Pierre-Antoine Bannier, Quentin Bertrand, Joseph Salmon, and 1 more author
    Medical imaging meets NeurIPS, 2021