Manuel Madeira

PhD student in Artificial Intelligence at EPFL

prof_pic.jpg

I am a PhD student at LTS4, at EPFL (Switzerland), under the supervision of Pascal Frossard and Dorina Thanou. My research focuses on the interplay between generative models, in particular diffusion, and data structure, often using graph generation as a testbed. I explore both how these models exploit inherent data structure and how to explicitly encode this structure into them, with applications to scientific discovery, such as molecular generation and digital pathology. I am also interested in extending these ideas to broader domains, e.g., language.

During my PhD, I’ve interned as an Applied Scientist at Amazon. Before the PhD, I was a Machine Learning researcher at Inductiva Research Labs, a startup where I worked on physics-informed neural networks for heat diffusion and coastal dynamics.

Prior to that, I completed my BSc and MSc in Biomedical Engineering from Instituto Superior Técnico. I was also a visiting student at Tsinghua University. For my MSc thesis, I worked on variance reduction for stochastic convex optimization with Renato Negrinho, Pedro Aguiar, and João Xavier.

In my spare time, I am really into sports! Back in Portugal, I used to play football at a competitive level. You can check my (Portuguese) football player page here and a young Manuel playing here (8 in red and black stripes). I do also enjoy great books and movies.

For a complete list of publications, check my Google Scholar page.

Selected Publications

  1. defog_thumbnail.png
    DeFoG: Discrete Flow Matching for Graph Generation
    Yiming Qin*, Manuel Madeira*, Dorina Thanou, and 1 more author
    arXiv, 2024
  2. construct_website_thumbnail.png
    Generative Modelling of Structurally Constrained Graphs
    Manuel Madeira, Clément Vignac, Dorina Thanou, and 1 more author
    NeurIPS, 2024
  3. tls_preview.png
    Tertiary Lymphoid Structures Generation through Graph-based Diffusion
    Manuel Madeira, Dorina Thanou, and Pascal Frossard
    GRAIL Workshop, MICCAI, 2023