Résultats de recherche

SynopsisA honeypot is a security tool deliberately designed to be vulnerable, thereby enticing attackers to probe, exploit, and compromise it. Since their introduction in the early 1990s, honeypots have remained among the most widely used tools for capturing cyberattacks, complementing traditional defenses such as firewalls and intrusion detection systems. They serve both as early warning systems and as sources of valuable attack data, enabling security professionals to study the techniques and behaviors of threat actors.While conventional honeypots have achieved significant success, they remain deterministic in their responses to attacks. This is where adaptive or intelligent honeypots come into play. An adaptive honeypot leverages Machine Learning techniques, such as Reinforcement Learning, to interact with attackers. These systems learn to take actions that can disrupt the normal execution flow of an attack, potentially forcing attackers to alter their techniques. As a result, attackers must find alternative routes or tools to achieve their objectives, ultimately leading to the collection of more attack data.Despite their advantages, traditional honeypots face two main challenges. First, emulation-based honeypots (also known as low- and medium-interaction honeypots) are increasingly susceptible to detection, which undermines their effectiveness in collecting meaningful attack data. Second, real-system-based honeypots (also known as high-interaction honeypots) pose security risks to the hosting organization if not properly isolated and protected. Since adaptive honeypots rely on the same underlying systems, they also inherit these challenges.This thesis investigates whether it is possible to design a honeypot system that mitigates these challenges while still fulfilling its primary objective of collecting attack data. To this end, it proposes a new abstract model for adaptive self-guarded honeypots, designed to balance attack data collection, detection evasion, and security preservation, ensuring that it does not pose a risk to the rest of the network.Membres du juryProf. Wim VANHOOF, Président, Université de NamurProf. Jean-Noël COLIN, Promoteur, Université de NamurProf. Florentin ROCHET, Membre interne, Université de NamurProf. Benoît FRENAY, Membre interne, Université de NamurProf. Ramin SADRE, Membre externe, Université catholique de Louvain Dr. Jérôme FRANCOIS, Membre externe, Université du LuxembourgVous êtes cordialement invités à un drink, qui suivra la soutenance publique. Pour une bonne organisation, merci de donner votre réponse pour le mardi 20 mai 2025.   Je m'inscris

SynopsisLes méthodes deep learning sont devenues de plus en plus populaires pour construire des systèmes intelligents. Actuellement, de nombreuses architectures deep learning constituent l'état de l'art dans leurs domaines respectifs, tels que la reconnaissance d'images, la génération de texte, la reconnaissance vocale, etc. La disponibilité de bibliothèques et de frameworks matures pour développer de tels systèmes est également un facteur clé de ce succès.Ce travail explore l'utilisation de ces architectures pour construire des systèmes intelligents pour les langues des signes. La création grands corpus de données en langue des signes a rendu possible l'entraînement d'architectures deep learning à partir de zéro. Les contributions présentées dans ce travail couvrent tous les aspects du développement d'un système intelligent basé sur l'apprentissage profond. Une première contribution est la création d’une base de données pour la Langue des Signes de Belgique Francophone (LSFB). Celle-ci est dérivé d’un corpus existant et a été adapté aux besoins des méthodes deep learning. La possibilité de recourir à des méthodes de collecte participative (crowdsourcing) pour recueillir d'avantages de données est également explorée.La deuxième contribution est le développement ou l’adaptation d'architectures pour la reconnaissance automatique de la langue des signes. L'utilisation de méthodes contrastives pour apprendre de meilleures représentations est explorée, et la transférabilité de ces représentations à d'autres langues des signes est évaluée.Enfin, la dernière contribution est l’intégration des modèles dans des logiciels destinés au grand public. Cela a permis de mener une réflexion sur les défis lié à l'intégration d'un module intelligent dans le cycle de vie du développement logiciel.Membres du juryProf. Wim VANHOOF, Président, Université de NamurProf. Benoît FRENAY, Promoteur, Université de NamurProf. Anthony CLEVE, Co-promoteur, Université de NamurProf. Laurence MEURANT, Membre interne, Université de NamurProf. Lorenzo BARALDI, Membre externe, Université de ModèneProf. Annelies BRAFFORT, Membre externe, Université de Paris-SaclayProf. Joni DAMBRE, Membre externe, Université de GandVous êtes cordialement invités à un drink, qui suivra la soutenance publique. Pour une bonne organisation, merci de donner votre réponse pour le vendredi 6 juin. Je m'inscris

SynopsisOver recent years, the development of agent-based models has allowed researchers to advance their understanding of naturally occurring collective behaviours. Swarm robotics, a field studying the design of decentralised robot swarms, has emerged following the replication of some collective behaviours in artificial groups of robots. The first part of this thesis provides novel techniques for the aggregation of heterogeneous swarms. First, we enhance an existing controller for an aggregation problem on two sites through the use of informed robots. We show that our simplified approach offers a wider range of operating conditions and a greater flexibility. Second, we provide a new method for the aggregation of robot swarms with adaptive random walks. We separately study cue-based aggregation with a swarm of robots only sensing private information and neighbour-based aggregation with a swarm of robots sensing social information. We show that a trade-off can be obtained with a heterogeneous swarm composed of the two robot types, forming a dense cluster near the minimum of an environmental cue. Private and social information also play a key role in the evolution of biological processes inside animal groups. Dispersal, the movement of an animal from site of birth to site of reproduction, is strongly affected by the acquisition and the use of information. Since experimental research is often difficult to conduct while accounting for multiple information sources and environmental variability, the use of agent-based models offer an opportunity to study the evolution of dispersal and its associated costs linked to private and social information in a controlled setting. The second part of this thesis provides an agent-based model of dispersal including the acquisition of information and its associated costs. Throughout three case studies, we observe the evolution of genes linked to the acquisition of information and the obtained dispersal strategies in different scenarios. Jury members Prof. Wim Vanhoof, Président, Université de Namur, BelgiqueProf. Elio Tuci, Secrétaire, Université de Namur, BelgiqueProf. Timoteo Carletti, Membre interne, Université de Namur, Belgique Prof. Eliseo Ferrante, Membre externe, Vrije Universiteit Amsterdam, Pays-BasProf. Mauro Birattari, Membre externe, ULB, Belgique Prof. Andreagiovanni Reina, Membre externe, Universität Konstanz, Allemagne  La défense sera suivie d'un drink. Je m'inscris

Abstract Radiotherapy is a cornerstone of cancer treatment and is currently administered to approximately half of all cancer patients. However, the cytotoxic effects of ionizing radiation on normal tissues represent a major limitation, as they restrict the dose that can be safely delivered to patients and, consequently, reduce the likelihood of effective tumor control. In this context, delivering radiation at ultra-high dose rates (UHDR, > 40 Gy/s) is gaining increasing attention due to its potential to spare healthy tissues surrounding the tumor and to prevent radiation-induced side effects, as compared to conventional dose rates (CONV, on the order of Gy/min).The mechanism underlying this protective effect—termed the FLASH effect—remains elusive, driving intensive research to elucidate the biological processes triggered by this type of irradiation.In vitro models offer a valuable tool to support this research, allowing for the efficient screening of various beam parameters and biological responses in a time- and cost-effective manner. In this study, multicellular tumor spheroids and normal cells were exposed to proton irradiation at UHDR to evaluate its effectiveness in controlling tumor growth and its cytotoxic impact on healthy tissues, respectively.We report that UHDR and CONV irradiation induced a comparable growth delay in 3D tumor spheroids, suggesting similar efficacy in tumor control. In normal cells, both dose rates induced similar levels of senescence; however, UHDR irradiation led to lower apoptosis induction at clinically relevant doses and early time points post-irradiation.Taken together, these findings further highlight the potential of UHDR irradiation to modulate the response of normal tissues while maintaining comparable tumor control.JuryProf. Thomas BALLIGAND (UNamur), PrésidentProf. Stéphane LUCAS (UNamur), SecrétaireProf. Carine MICHIELS (UNamur)Dr Sébastien PENNINCKX (Hôpital Universitaire de Bruxelles)Prof. Cristian FERNANDEZ (Université de Bern)Dr Rudi LABARBE (IBA)

Abstract Due to their unique chemical, physical and photophysical properties, organoboron compounds and in particular triarylboranes play a central role in chemistry and in catalysis. Trivalent neutral boron Lewis acids, which are planar trigonal species, have been shown to exhibit enhanced Lewis acidity and electrophilicities when constrained in a pyramidal trigonal environment. Within the context of the emerging area of geometrically constrained main-group elements, the fundamental experimental and computational investigations of the impact of structural deformation on the physicochemical properties and reactivity of borane derivatives is of interest. This thesis will explore successively the development of geometrically constrained intramolecular FLP and of cationic boron Lewis superacid based on the aza-boratriptycene scaffold, then the synthesis of pyramidalyzed electron-deficient borenium cation with tethered pyridine and NHC ligands embedded in the triptycene scaffold and will finally focus on chiral borenium cations as new Lewis acids. A collaborative work dealing with the combination of the strong 9-sulfonium-10-boratriptycene with hindered Lewis bases is finally performed for developing latent FLP. This work deepens our understanding of the synthesis of constrained boron Lewis acids species, a key step to develop new pyramidal boron Lewis superacids, deblocking new kinds of reactivity in main-group chemistry. For instance, electrophilic Csp2–H borylation reactions of electron-poor aromatics were observed, new unusual binding mode at weakly coordinating anions were discovered and encouraging steps were initiated for reaching new chiral boron-based Lewis acids, opening the path toward new horizons in main-group chemistry.JuryProf. Benoît CHAMPAGNE (UNamur), PrésidentProf. Guillaume BERIONNI (UNamur), SecrétaireProf. Olivier CHUZEL (Aix-Marseille Université)Prof. Raphaël ROBIETTE (UCLouvain)Prof. Stéphane VINCENT (UNamur)

En mai 2025, le Département de physique recevait des visiteurs particuliers : deux namurois, Serge Mathot et François Briard, alumni de l’UNamur et membres du CERN.  Plusieurs activités étaient au programme, allant de la visite de l’accélérateur à particules, en passant par la vulgarisation scientifique et les séminaires thématiques notamment en sciences du patrimoine.  Objectif ? Identifier les domaines ou activités dans lesquels l’UNamur et le CERN pourraient renforcer leur collaboration.