Semantics Scholar / Github / Email

I am currently a PhD candidate at CMAP (Ecole Polytechnique) and the Institut Liryc, where I work with Eric Moulines and Rémi Dubois. Prior to this, I was a Machine Learning Engineer at Volta Medical.

My PhD research focuses on Generative Models, with a particular interest in applications to Cardiology and general health data. Recently, I have been working with diffusion models and their usage as a prior for solving ill-posed inverse problems.

---

Talks and Resources

• Introduction to Score Based Generative models presented in the Hi! Paris reading group. (Link for notebook in the github)

  HTML version of the notebook and Slides (PDF)

Research

• Diffusion posterior sampling for simulation-based inference in tall data settings

  J. Linhart, G. Cardoso, A. Gramfort, S. Le Corff and P. Rodrigues.
  Under review.
  

• Bayesian ECG reconstruction using denoising diffusion generative models

  G. Cardoso, L. Bedin, J. Duchateau, R. Dubois and E.Moulines.
  Under review.
  

• Monte Carlo guided Diffusion for Bayesian linear inverse problems

  G. Cardoso, Y. Janati, S. Le Corff and E. Moulines.
  International Conference on Learning Representations (ICLR) 2024. Oral presentation
  

• State and parameter learning with PaRISian Particle Gibbs

  G. Cardoso, YJEL, S. Le Corff, E. Moulines and J. Olsson.
  International Conference on Machine Learning (ICML) 2023.
  

• Particle-based, Rapid Incremental Smoother Meets Particle Gibbs

  G. Cardoso, E. Moulines and J. Olsson.
  Statistica Sinica (2023).
  

• BR-SNIS: Bias Reduced Self-Normalized Importance Sampling

  G. Cardoso, S. Samsonov, A.Thin, E. Moulines and J. Olsson.
  Neural Information Processing Systems(NeurIPS) 2022.
  

• A Patient-Specific Single Equivalent Dipole Model

  G. Cardoso, G. Robin, A. Arrieula, M. Potse, M. Haïssaguerre, E.Moulines and R. Dubois.
  2022 Computing in Cardiology (CinC) 2022.
  

• Generative methods for sampling transition paths in molecular dynamics

  T. Lelièvre, G. Robin, I. Sekkat, G. Stoltz, G. Cardoso.
  ESAIM Proceedings and Surveys 2022.