Search results

At the beginning of each academic year, the Board of Trustees grants Namur Research College (NARC) Fellowship status to researchers who demonstrate a high level of research achievement and who have recently received a prestigious award or funding. A look back at the fellowship of Professor Frederik De Laender.

In 2021, the European Union has embarked on a titanic project to safeguard the genomes of all eukaryotic species in Europe. In other words, all living organisms, with the exception of bacteria and archaea (micro-organisms). Called ERGA, for European Reference Genome Atlas, and in which UNamur is participating thanks to Professor Alice Dennis, this project hopes to help safeguard biodiversity, at a time when a fifth of European species are in danger of extinction..

The F.R.S.-FNRS has just published the results of its various 2024 calls. Equipment calls, research credits and projects, FRIA doctoral grants and Mandant d'Impulsion Scientifique (MIS), there are many instruments to support fundamental research. Find out more about UNamur's results.

Jamie Wright, a researcher at the Namur Digital Institute at UNamur, participated in a study conducted by Pompeu Fabra University (UPF) in Barcelona and published in the prestigious journal *Nature Human Behaviour*.  The study shows that languages around the world tend towards lexical economy, reusing words to denote different concepts when doing so does not lead to communicative confusion.

The result of a collaboration with Isabelle Frochot of the University of Burgundy Europe, a study in which Alain Decrop, a marketing professor at the EMCP Faculty, participated, shows that the mountains foster a genuine reconnection with oneself, with others, and with one’s environment. The results of this research were published in the prestigious journal *Recherche et Applications en Marketing* (RAM).

Through research, teaching, and knowledge sharing, Wafa Hammedi, a professor at Sciences Po’s Faculty of Economics, Management, and Communication (EMCP), has charted a path guided by her convictions. Far from the performance-driven clichés typically associated with marketing, she advocates for socially engaged research that serves society, equity, and inclusion.

Abstract Deep learning has become a central part of everyday life, and has given us powerful digital tools such as machine translators, voice assistants, and large language models. The Deaf communities, however, benefit verylittle from this progress. Sign languages are full natural languages, but they have no widely used written form, so the large text corpora and pretrained models that support these tools for spoken languages have no direct equivalent. As a result, sign language processing is held back by two related problems: a persistent shortage of annotated data, and a lack of tools to analyze the languages themselves.This thesis focuses on one component of the sign language processing pipeline that has received comparatively little attention: sign language segmentation, the temporal partitioning of a continuous signing video into individual sign units. Segmentation sits between raw video and almost every symbolic task built on top of it, yet it remains poorly understood. The goal of this work is twofold: to advance the segmentation task itself, and to turn sign language processing research into tools that are usable by the Deaf community, with French Belgian Sign Language (LSFB) as the main setting.The central contribution of this thesis is a set of segmentation models. We first study approaches based on recurrent neural networks, and show that their main difficulty lies in the modeling of the transitions between signs, the brief ambiguous movements that separate one sign from the next and that are easily confused with the signs themselves. This insight motivates the Hydra framework, a new approach that detects each sign as a whole unit rather than deciding, frame by frame, whether a sign is being produced. Concretely, instead of classifying every frame, the model predicts the distance from each frame to the nearest sign boundaries, which makes it less sensitive to the ambiguous regions and improves consistently over prior methods. Around this core, we present further contributions: a sign language-to-text dictionary that recognizes signs from an ordinary webcam, supporting work on isolated sign recognition, and a collaborative platform that gathers new sign data as a side effect of everyday use. Together, these contributions narrow the gap between sign language processing research and the people whose language it concerns. Jury Prof. Benoit Frénay - University of Namur, BelgiumProf. Anthony Cleve - University of Namur, BelgiumProf. Katrien Beuls - University of Namur, BelgiumProf. Barbara Hammer - Bielefeld University, GermanyProf. Michèle Gouiffès - Paris-Saclay University, FranceDr. Mathieu De Coster - Ghent University, Belgium I am registering to attend Pierre Poitier's thesis defense

Catherine Linard, a researcher in the Department of Geography at UNamur, was recognized for her work at the intersection of geography and epidemiology. This Francqui-Collen fellowship will allow her, starting in September and for the next three years, to be relieved of most of her teaching duties so she can devote more time to her research on the influence of environmental changes on the spread of infectious diseases. 

Researchers at UNamur have developed a new method that has uncovered 32 previously undetected bugs in some of the most widely used software libraries for training artificial intelligence models. Originating from a master’s thesis, this research has already led to fixes in several software programs used by millions of developers around the world. 

AbstractIn this work, I aimed to accurately simulate the Raman optical activity (ROA) signatures of flexible solute molecules in their environment. Indeed, these systems are quite challenging due to the potentially huge number of conformers—that is, local minima on the potential energy surface (PES). Furthermore, the environment can drastically alter the PES, as evidenced by experimental ROA spectra that vary significantly depending on the molecule’s conformation and its surroundings. To address this issue, we have developed a hierarchy of methodologies designated M1, M2, and M3. In M1 and M2, the so-called static approach is used: one or more conformers are identified, their geometries are reoptimized at the Density Functional Theory (DFT) level, and their vibrational signatures are simulated, typically using the harmonic approximation. The initial list of conformers was obtained using the CREST algorithm. In M1, the solvent is treated implicitly, whereas in M2, explicit solvent molecules are added around the solute molecule. In M3, we perform ab initio molecular dynamics simulations of the solute molecule surrounded by explicit solvent molecules. The spectrum is obtained by evaluating time-correlation functions along the trajectory. Our first systems of interest were cryptophane derivatives. These are flexible cage-like systems consisting of two hemispheres connected by three -O-(CH₂)n-O- (denoted Cr–nnn) linkers that exhibit chiroptical properties. For Cr–111, the smallest possible cryptophane, the M1 methodology showed good agreement with experimental data, especially in the fingerprint region. However, we demonstrated that the relative ratio between the different conformers was strongly influenced by the choice of the exchange-correlation (XC) functional in our DFT calculations, highlighting the sensitivity of the potential energy surface (PES) description. Cr–222 molecules have been shown to be more flexible, as evidenced by a greater number of significant conformers. The overall agreement with experimental data was also satisfactory. Finally, to test our different approaches (M1–M3), I performed new ROA measurements of amino acids in water at the University of Bordeaux in Dr. Daugey’s laboratory. When compared to our simulations, we clearly observed an improvement in the ROA signatures when explicit water molecules were added to our simulations (M2 vs. M1). Unfortunately, the M3 method did not perform as expected, and further investigation is needed. Overall, I have shown that the PES, as described by our various methodologies, is highly sensitive to various simulation parameters—such as the XC functional, the number and position of explicit solvent molecules, and so on—and that all of these factors strongly influence the simulated ROA signatures.JuryProf. Francesca CECCHET (UNamur), ChairProf. Vincent LIÉGEOIS (UNamur), SecretaryProf. Benoît CHAMPAGNE (UNamur)Prof. Carine CLAVAGUÉRA (University of Paris-Saclay)Dr. Nicolas DAUGEY (University of Bordeaux)

AbstractEnvironmental change rarely occurs as a single, isolated event; organisms and communities typically experience one stressor against the backdrop of another that has already passed. Arunima Sikder’s doctoral thesis examines whether an organism’s past environment shapes its response to the environment that follows, and whether that influence propagates upward to affect the stability of the communities to which these organisms belong.The thesis addresses this question in Synechococcus sp., a globally significant marine primary producer, using thermal and chemical stressors as its two environmental drivers. Working across three levels of biological organization, the study examines how past environments shape responses to subsequent ones—whether they occur in sequence, fluctuate, or occur in combination—using functional traits as the common metric for measuring these responses.The results reveal three key findings. First, sensitivity to a stressor depends on the sequence of exposure, not on the stressor alone. Second, response diversity predicts community stability only when measured under the community’s actual acclimation history, rather than under sustained conditions as conventional wisdom assumes. Third, an apparent contradiction in biodiversity–ecosystem function theory—a negative relationship between functional diversity and performance.Together, these findings suggest that sensitivity, response diversity, and functional diversity should be treated as state-dependent quantities rather than fixed properties, with implications for how the performance of biological systems is measured and modeled under increasingly variable environmental conditions.JuryProf. Alice DENNIS (UNamur), ChairProf. Frederik DE LAENDER (UNamur), SecretaryProf. Maren STRIEBEL (Carl von Ossietzky University Oldenburg)Prof. Giulia GHEDINI (Monash University)Prof. Eli THORÉ (UNamur)

The days when people chose a career once and for all and when experience alone was enough to climb the career ladder seem to be over. Nowadays, lifelong learning has almost become the norm. “There is clearly a trend in this direction, particularly in the context of automation and artificial intelligence, which makes acquiring new skills all the more important,” says Michel Ajzen, a professor in the Management Department at the Faculty of Economics, Management, and Communication at SciencePo, UNamur, and a member of the Namur Digital Institute. “And it’s not a generational issue: whether you’re 30, 40, or 50, this need exists.”