prof_pic.jpg

Thibaut Issenhuth

I am a researcher in deep generative models at Criteo AI Lab. I completed my PhD in 2023 in collaboration with Criteo AI Lab and Imagine Lab, Ecole des Ponts ParisTech. I was advised by Jérémie Mary (Criteo) and David Picard (ENPC).

Research interest. My research focuses on the use of deep generative models on high-dimensional data, particularly on natural images. I am interested in both the theoretical understanding and the potential applications of these models. My main contributions concern improving and understanding Generative Adversarial Networks (GANs) on multi-modal and disconnected data distributions, and deep neural networks for image editing.

selected publications

  1. NeurIPS
    Unifying GANs and Score-Based Diffusion as Generative Particle Models
    Jean-Yves Franceschi, Mike Gartrell, Ludovic Dos Santos, and 4 more authors
    In Conference on Neural Information Processing Systems 2023
  2. ICML
    Unveiling the latent space geometry of push-forward generative models
    Thibaut Issenhuth, Ugo Tanielian, Jérémie Mary, and 1 more author
    In International Conference on Machine Learning, 2023
  3. TMLR
    EdiBERT: a generative model for image editing
    Thibaut Issenhuth, Ugo Tanielian, Jérémie Mary, and 1 more author
    Transactions on Machine Learning Research, 2022
  4. WACV
    Latent reweighting, an almost free improvement for GANs
    Thibaut Issenhuth, Ugo Tanielian, David Picard, and 1 more author
    In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2022
  5. ICML
    Learning disconnected manifolds: a no gan’s land
    Ugo Tanielian, Thibaut Issenhuth, Elvis Dohmatob, and 1 more author
    In International Conference on Machine Learning, 2020
  6. ECCV
    Do not mask what you do not need to mask: a parser-free virtual try-on
    Thibaut Issenhuth, Jérémie Mary, and Clément Calauzènes
    In European Conference on Computer Vision, 2020
  7. IJCARS
    Face detection in the operating room: Comparison of state-of-the-art methods and a self-supervised approach
    Thibaut Issenhuth, Vinkle Srivastav, Afshin Gangi, and 1 more author
    International journal of computer assisted radiology and surgery, 2019