Vous êtes ici : Accueil » Kiosque » Annonce


Mot de passe : 

Mot de passe oublié ?
Détails d'identification oubliés ?


14 avril 2015

PHD: Physiological signal fusion for affective computing

Catégorie : Doctorant

We propose a PhD subject dealing with the fusion of information enclosed in bio-signals in order to evaluate the nervosity of a patient.

All information for candidacy avaible at this link. Further information on the PhD available at this link.


Following new trends in connected devices (smart watches, connected bracelet), several physiological signals can be acquired. These bio-signals are in general related to cardiac activity, sweating or muscular activity. Many applications (especially for health care) are arising and companies in this business line have remarkable growth rates. When signals are processed efficiently, it is possible for instance to characterize the nervosity of a patient. For individuals suffering from autism such an emotional state is a feature of the premises of a violent crisis. Another potential application deals with elder people. Equipped with a connected bracelet, an elder person can be taken care of quickly in case of fall or panic attack.

Signal processing in this framework requires advanced theoretical developments. In order to address bottle-neck problems in this field, we propose to PhD subject in applied mathematics for signal processing. The future PhD student will have to design new statistical models that take into account all signal imperfections such as uncertainty in measures and imprecision in parameters. Probability theory is an efficient framework to mitigate uncertainty. When, on top of that, parameters are ill-known, more general formalisms are useful. This formalisms mainly rely on Choquet capacities which are more general than probability measures. Roughly speaking, a capacity may be viewed as the upper (or lower) bound of a set probability measures.

This PhD follows a recent collaboration in affective computing with Laurent Sparrow who is an associate professor at Lille3 University and a member of SCALab UMR 9193.

The ideal candidate should have a master degree in applied mathematics (or any equivalent engineer degree) with some experience in signal processing. The future PhD student will be part of the SIGMA team of laboratory CRIStAL UMR CNRS 9189.


Dans cette rubrique

(c) GdR 720 ISIS - CNRS - 2011-2018.