Establishing guarantees for crafting and defending against adversarial attacks.
Neural networksProximal algorithms
Towards the unification of deep neural networks and iterative proximal methods.
Scale invariance analysis of fetal heart rate.
MultifractalFHRFetal AcidosisSparse SVM
This project aims at optimizing the hyperparameters of models learned by minimizing nonsmooth functionals
Bilevel optimizationNonsmoothBregman
This project deals with the case where the scale invariance properties are inhomogenous in time or in space.
MultifractalSegmentationTextures
Joint analysis of the scale invariance of multivariate data.
MonofractalMultivariate dataBranch and bound algorithmInternet traffic
Learning physical problems and incorporating/unveiling background knowledge
Neural operatorsPhysics-informed modelsDynamical systems
Various contributions about piecewise constant denoising.
Potts modelTotal-variationOnline denoisingRegularization parameterMultivariate denoising