Marius Dubosc

PhD Student

Team

Image Processing and Pattern Recognition

Campus

Paris

Short Bio

I have been working for 2 years on Doppler Holography, a new medical imaging method. After contributing to the acquisition software in C++/CUDA, I started a PhD on deep learning applied to this imaging technique, under the management of Thierry Géraud, Edwin Carlinet, Nicolas Boutry and Michael Atlan. Retinal vessel segmentation, finding biomarkers through machine learning and data analysis or even classification of pathologies are among the project.

Previously, I obtained an Engineering Degree in Computer Science from EPITA in 2023 with a double major in datascience and artificial intelligence, and scientific research.

Research

My main research interests are:

  • Computer vision
  • Medical imaging
  • Frugal machine learning (compression of neural networks)
  • Doppler Holography