Publications

Machine Learning

Correcting Neural Operator Spectral Bias via Diffusion Posterior Sampling with Sparse Observations

Neural operator surrogates (NO) approximate PDE solutions orders of magnitude faster than numerical solvers, but suffer from spectral bias: high-frequency content is systematically attenuated, limiting reliability where fine-scale structure matters…

Niccolò Perrone, Fanny Lehmann, Stefania Fresca, Filippo Gatti

Materials and structures in mechanics

ReaderAdaptNet: modeling reader variability in breast imaging with reader-specific embeddings

Abstract Objective. Inter-reader variability remains a major challenge in breast imaging interpretation, particularly for ordinal classification tasks such as breast density and background parenchymal enhancement (BPE). These visual assessments are…

Elodie Ripaud, Clément Jailin, Pablo Milioni de Carvalho, Laurence Vancamberg, Isabelle Bloch

Physics in Medicine and Biology - 2026

Engineering Sciences

Mechanics-guided data-enhanced kinematic models for time-resolved DVC

X-ray Computed Tomography (XCT) combined with Digital Volume Correlation (DVC) enables for internal displacement and strain measurements in deforming materials. The long acquisition time of tomographic scans restricts analyses to a few static…

Pierre Latil, Malo Valmalle, Benjamin Smaniotto, Clément Jailin, François Hild

Comptes Rendus. Mécanique - 2026

Instrumentation and Detectors

Integrated Projection-enhanced DVC to monitor nonlinearities associated with fracture experiments

Digital Volume Correlation (DVC) is an established method for extracting full-field displacements from in-situ experiments monitored by X-ray computed tomography. However, the temporal sampling of such measurements is usually sparse, which limits…

Malo Valmalle, Clément Jailin, François Hild

17ème Colloque National en Calcul des Structures (CSMA 2026) - 2026

Structural mechanics

Hybridation de modèles pour la résolution de problèmes thermo- poroélastiques non-linéaires fortement couplés

Elise Foulatier, Pierre-Alain Boucard, François Louf, David Néron

17ème Colloque National en Calcul des Structures - 2026

Engineering Sciences

Damage of a structure under multiaxial dynamic inertial loading: A lumped damage approach

Extracting model parameters from experiments mimicking earthquake loading conditions is of interest to better design civil engineering structures. In the present case, the challenge is to identify the complex response of a notched section of a…

Xuyang Chang, Jan Grashorn, Marius Bittner, François Hild, Stéphane Roux

European Journal of Mechanics - A/Solids - 2026

Engineering Sciences

Continuous structural health monitoring with a modified dual Kalman filter applied with optical fiber sensing data

The advent of new sensing technologies has led to an increased use of sensors for the monitoring of systems and structures. By integrating data from these sensors with numerical models, a customized digital twin of the struc ture can be constructed…

S. Farahbakhsh, Ludovic Chamoin, M. Poncelet

Computers & Structures - 2026

Structural mechanics