Publications

• Poster de conférence

Identification de défauts pour le monitoring des rails

Elissa Akiki, Lynda Chehami, Nikolay Smagin, Emmanuel Moulin, Jamal Assaad, Youssef Zaatar

L’évaluation non destructive des rails est cruciale pour garantir une sécurité élevée dans le transport ferroviaire. On propose ici une méthode d’identification des sources de bruit et/ou défauts par apprentissage sur une base de données de référence. Des essais expérimentaux ont été effectués sur…, Non-destructive evaluation of rails is crucial to ensure high safety in rail transport. Here we propose a method for identifying sources of noise and/or defects by learning on a reference database. Experimental tests were carried out on a rail subjected to a force torque applied at different…

Mardi des Chercheurs 2024, Mar 2024, Mons, Belgique. ⟨hal-04542488⟩

• Poster de conférence

Localisation 3D via des réseaux de capteurs en milieu électromagnétiquement perturbé: contexte de l’usine 4.0

Thomas Steve NKOLO BEKONO, Michael Bocquet, S. Baranowski

L’usine connectée, ou usine du futur, va utiliser les équipements déjà existants, les améliorer, puis les placer au cœur d’un réseau où toutes les informations pourront être échangées. Cela va mettre en évidence l’intérêt : de l’internet des objets, des robots collaboratifs (cobots), de la réalité…, The connected factory, or factory of the future, will use existing equipment, improve it and then place it at the heart of a network where all the information can be exchanged. This will highlight the benefits of the Internet of Things, collaborative robots (cobots), augmented reality, cyber-…

Mardi des chercheurs, Mar 2024, Mons, Belgique. ⟨hal-04524864⟩

• Poster de conférence

Leveraging Efficient Models for Recognizing Drivers' Facial Expressions

Ibtissam Saadi, Douglas W Cunningham, Abdelmalik Taleb-Ahmed, Abdenour Hadid, Yassin El Hillali

Our new approach, ShuffViT-DFER, combines lightweight CNN and vision transformer models for efficient and accurate real-time driver facial expression recognition. By merging their features effectively, we outperform existing methods on benchmark datasets like KMU-FED and KDEF.

Mardi des chercheurs 2024, Mar 2024, mons, Belgium. ⟨hal-04546700⟩

• Article dans une revue

EME-CNTK: Infinite Limits of Convolutional Neural Network for Urban Electromagnetic Field Exposure Reconstruction

M. Mallik, B. Allaert, E. Egea-Lopez, D.P. Gaillot, J. Wiart, L. Clavier

Electromagnetic field (EMF) exposure has grown to be a critical concern as a consequence of the ongoing installation of fifth-generation cellular networks (5G). The lack of measurements makes it difficult to accurately assess the EMF in a specific urban area, as Spectrum cartography (SC) relies on…

IEEE Access, 2024, 12, pp.49476 - 49488. ⟨10.1109/ACCESS.2024.3380835⟩. ⟨hal-04537011⟩

• Article dans une revue

Design and testing of a multi-electrode apodized acousto-optic filter for arbitrary polarized light

Samuel Dupont, Jean-Claude Kastelik, Jurgen Vanhamel

We present an original acousto-optic tunable filter that is able to filter visible light from 400 to 650 nm and is designed to interact simultaneously with two polarizations. The filter shows an adjustable optical bandwidth and apodization capabilities. These features make it suitable for practical…

Applied optics, 2024, 63 (10), pp.2487. ⟨10.1364/AO.514229⟩. ⟨hal-04516407⟩

• Article dans une revue

Boosting Gas-Phase TiO 2 Photocatalysis with Weak Electric Field Strengths of Volt/Centimeter

My Nghe Tran, Myriam Moreau, Ahmed Addad, Adrien Teurtrie, Thomas Roland, Vincent de Waele, Marc Dewitte, Louis Thomas, Gaëtan Lévêque, Chunyang Dong, Pardis Simon, Karima Ben Tayeb, D. Mele, Vitaly Ordomsky, Bruno Grandidier

Among semiconductor nanomaterials, titanium dioxide is at the forefront of heterogeneous photocatalysis, but its catalytic activity greatly suffers from the loss of photoexcited charge carriers through deleterious recombination processes. Here, we investigate the impact of an external electric…

ACS Applied Materials & Interfaces, 2024, 16 (12), pp.14852-14863. ⟨10.1021/acsami.3c19031⟩. ⟨hal-04526258⟩

• Poster de conférence

Thermo-mechanical influence on fracture propagation: integrating temperature effects through equilibrium statistical mechanics

Claudia Binetti, Giuseppe Florio, Nicola Pugno, Stefano Giordano, Giuseppe Puglisi

Rencontre du non-linéaire 2024, Mar 2024, Paris, France. . ⟨hal-04508600⟩

• Communication dans un congrès

Design of Microstrip UWB Antenna With Full Ground Plane For Wearable Applications

Chukwuka Ozuem Anthony, Divitha Seetharamdoo, Florent Gamand, Christophe Gaquiere

2024 18th European Conference on Antennas and Propagation (EuCAP), Mar 2024, Glasgow, France. pp.1-4, ⟨10.23919/EuCAP60739.2024.10501615⟩. ⟨hal-04575406⟩

• Article dans une revue

Inverse Design of One-Dimensional Topological Photonic Systems Using Deep Learning

M. El Ghafiani, M. Elaouni, S. Khattou, Y. Rezzouk, M. Amrani, Othmane Marbouh, Mohamed Boutghatin, Abdelkrim Talbi, E. El Boudouti, B. Djafari Rouhani

We demonstrate a novel approach to inversely design one-dimensional (1D) photonic stubbed systems with targeted topological properties by leveraging the power of deep learning. The process involves developing a data-driven model to accurately predict the geometric parameters of the photonic system…

Physics of Wave Phenomena, 2024, 32 (1), pp.48-55. ⟨10.3103/S1541308X24010047⟩. ⟨hal-04511253⟩

• Article dans une revue

Emb-trattunet: a novel edge loss function and transformer-CNN architecture for multi-classes pneumonia infection segmentation in low annotation regimes

Fares Bougourzi, Fadi Dornaika, Amir Nakib, Abdelmalik Taleb-Ahmed

Abstract One of the primary challenges in applying deep learning approaches to medical imaging is the limited availability of data due to various factors. These factors include concerns about data privacy and the requirement for expert radiologists to perform the time-consuming and labor-intensive…

Artificial Intelligence Review, 2024, 57 (4), pp.90. ⟨10.1007/s10462-024-10717-2⟩. ⟨hal-04520691⟩