Matthieu Dabrowski
- Université de Lille
- Centre for Digital Systems (CERI SN - IMT Nord Europe)
Présentation
My name is Matthieu Dabrowski, I am a french post-doctoral researcher in AI and Computer Vision. I am mainly interested in projects related to Health or Environmental issues. I have developed an expertise in Physics-Informed Neural Networks. During my thesis this concept was applied to atmospheric physics. I am currently working on biomechanics-informed models for human pose prediction from feet pressure data. I am currently open for work and looking for a postdoctoral researcher position, starting from May 15th.
Domaines de recherche
Compétences
Publications
Publications
|
|
Knowledge-inspired fusion strategies for the inference of PM 2.5 values with a neural networkGeoscientific Model Development, 2025, 18 (12), pp.3707-3733. ⟨10.5194/gmd-18-3707-2025⟩ |
HUMAPS-4D : A Multimodal Dataset for HUman Motion Analysis with Physiological and Semantic informationsThe IEEE/CVF Conference on Computer Vision and Pattern Recognition 2026, Jun 2026, Denver (Colorado), United States |
|
Semi-supervised GAN with sparse ground truth as Boundary ConditionsIEEE International Joint Conference on Neural Networks 2023, Jun 2023, Queensland, Australia |
|
|
Méthodes d'apprentissage profond inspirées par la physique pour l'inférence de la qualité de l'airApprentissage [cs.LG]. Université de Lille, 2024. Français. ⟨NNT : 2024ULILB043⟩ |