Projects

I am currently involved in the following projects

• Revert Project. This project is funded by the European Interreg program France-Angleterre and regroups a consortium of hospitals and universities (Cambridge, UPJV and Artois). Its goal is to improve the treatment of Normal Pressure Hydrocephalus, a degenerative disease of the brain that can be treated by means of shunts. Our team at the LML is in charge of developing machine learning models that help clinicians on assessing the chances of positive outcomes of shunting operations.

• France  Relance project with Median Technologies. It is funded by the "Plan France Relance".  We collaborate on applications of self-supervised learning to the medical domain (segmentation of 3D CT-scans).

• APE project ("Apprentissage Profond des Emotions). In this collaboration (labcom) with the company FIPSICO, we apply deep learning to the problem of remote photo-plethysmography. It is funded by the EU and the Région Nord.

• Vivah. A joint Ph.D. project funded by the ANR. The aim of this collaboration with the CRIL and the UCCS is to apply deep learning and more precisely Graphical Neural Networks to the problem of discovery of new materials.

• MAIA project. This project enables to fund the Ph.D. research of Benoît Brebion for applications of self-supervised learning methods to time series in medical domain. It takes place in a collaboration with the team or Fabrice Wallois (UPJV, EEG and MEG for foetus and newborns), and the team of Olivier Balédent (analysis of cerebrospinal fluid and of blood in the brain for diagnosis of Normal Pressure Hydrocaephalus). 

• MOSOPS project with the Faire Faces institute (CHU Amiens and UPJV) and the LIM (UPJV). The aim of this project is to monitor the mobility of a face after surgery by means of Mocap. Machine learning models are used to generate from a video a mesh corresponding to a face.  Its vertices can then be used to assess the mobility of the face.