1st Early-Career Researchers’ Day of the DATA Programme I-SITE CAP 20-25

Liste des Contributions

11. Jet Classification with Particle Transformers: A Multiclass Learning Approach

26/02/2026 10:50

Contribution orale

Applied Machine Learning

In high-energy collisions, jets, which are collimated sprays of particles, can originate from various fundamental particles, including W and Z bosons, top quarks, and the Higgs boson. Accurately identifying these jets is crucial for studying Standard Model processes and investigating new physics beyond its framework. This study, conducted within the ATLAS collaboration at the Large Hadron...

8. OnlineBootKNN: An Unsupervised Framework for Detecting Anomalies in Spectral Data Streams

26/02/2026 11:10

Contribution orale

Anomaly Detection

Field: AI, Affiliation: UCA

Monitoring the elemental composition of materials in order to detect abnormal conditions in real-time is essential for applications like manufacturing quality control, environmental monitoring, and space exploration. This is achieved using sensors that analyze the interaction of a material with electromagnetic radiation, producing spectral data streams or a...

20. Characterization and anomaly detection in daily cow activities using wavelet-based features

26/02/2026 11:30

Contribution orale

Anomaly Detection

Anomaly detection in the day-to-day activity of dairy cows
is challenging, as true abnormal behavior must be distinguished from
the individual variability and the animals’ endogenous rhythms. Current
algorithms for anomaly detection in times series include various tech
niques, with neural network-based methods being the most prominent.
However, these approaches lack interpretability,...

25. Development of Statistical and Predictive Models for Anomaly Detection and Quality Control Improvement Based on Heterogeneous Industrial Data in Medical Device Manufacturing

26/02/2026 14:00

Contribution orale

Anomaly Detection

This postdoctoral research is conducted within the Laboratory of Computer Science, Modeling and Optimization of Systems (LIMOS) and falls within one of its research themes, Data, Services and Intelligence (DSI), in close alignment with the activities of the Department of Mathematical and Industrial Engineering. It focuses on the optimization of quality control processes in medical textile...

13. Apport des motifs gradual dans pour l'IA explicable

26/02/2026 14:20

Contribution orale

Foundations and Sustainability in Machine Learning

Les modèles ensemblistes et les réseaux de neurones profonds démontrent de très bons résultats dans les tâches de classification. Cependant, leur nature « boîte noire » empêche leur déploiement généralisé dans des domaines critiques comme la santé. L'IA explicable vise à rendre ces modèles plus compréhensibles. Dans la littérature, les résultats de classification sont expliqués principalement...

22. Fluctuations and concentration in two-layer neural networks

26/02/2026 14:40

Contribution orale

Foundations and Sustainability in Machine Learning

We study the learning dynamics of wide two-layer neural networks trained by stochastic gradient descent (SGD), aiming to understand quantitatively how network width shapes both the typical training trajectory and the variability of the final predictor.

We adopt an interacting particle viewpoint in which neurons evolve under SGD as a large coupled system. As the width grows, this...

9. Contributing to reproducible software measurements of energy consumption for machine learning

26/02/2026 15:00

Contribution orale

Foundations and Sustainability in Machine Learning

This work is part of the AI domain of the DATA program, and more specifically about sensors to obtain the energy consumption and machine learning algorithms. This work is being carried out in the LIMOS laboratory.
High-performance computing require careful management of power and energy budgets. Much of the work to achieve these energy-consumption goals will have to be done through hardware...

7. Towards Sustainable DBMS: A Framework for Real-Time Energy Estimation and Query Categorization

26/02/2026 15:40

Contribution orale

Foundations and Sustainability in Machine Learning

Abstract. Energy efficiency in database management systems (DBMS) is increasingly critical due to the rising computational demands of modern applica-tions. Our work proposes a complete framework to analyze energy consumption. We developed a real-time monitoring framework that captures CPU and memory utilization during query execution and estimates energy consumption. We have implemented a...

17. Stylistic-STORM : Désenchevêtrement spectral auto-supervisé Utilisation de l’apprentissage antagoniste et JEPA pour l’analyse météorologique

26/02/2026 16:00

Contribution orale

Applied Machine Learning

La conduite autonome exige des véhicules qu’ils perçoivent leur environnement (véhicules, piétons, feux, etc.) et qu’ils restent fiables sous des conditions changeantes. Lors d’un trajet, par exemple de Clermont-Ferrand à Paris, la météo peut basculer rapidement d’un temps clair à une pluie, du brouillard ou une neige intense. Ces conditions dégradées ne réduisent pas seulement la visibilité :...

18. A simple self-supervised model for modeling human visual learning

26/02/2026 16:20

Contribution orale

Applied Machine Learning

Children quickly develop powerful visual representations that support visual recognition, such as object recognition, with minimal supervision. However, the principles underpinning this development are poorly understood. In particular, it is unclear how natural experience interacts with unsupervised learning mechanisms to shape semantic representations.

This study explores whether the...

15. Adaptive Real-time Feedback for Group Discussions using Reinforcement Learning

Poster

Poster Flash Talks

Real-time feedback has the potential to significantly improve group dynamics in collaborative environments. However, delivering timely and context-aware interventions remains a challenge especially in the absence of annotated data. In this work we explore the use of reinforcement learning (RL) to provide real-time, adaptive feedback at the group level that promotes effective group...

14. Automatic Deep Deformable Registration using Domain Adaptation and Run-Time Optimisation

Poster

Poster Flash Talks

Augmented reality from preoperative 3D model registration is promising to assist navigation in minimally-invasive liver surgery. The current registration methods are either accurate, but require surgeon interactions to annotate anatomical landmarks, or are fully automatic, but inaccurate. We propose a two-step automatic and accurate registration method. Step 1) segments the registration...

23. Concept-Based Learning for Explainable AI

Poster

Poster Flash Talks

Field: Machine Learning / Explainable Artificial Intelligence
Affiliation: LIMOS

Deep learning models have achieved remarkable performance across a wide range of tasks, yet their lack of transparency remains a major obstacle to deployment in high-stakes and human-centered settings. My research addresses this challenge through concept-based learning, a paradigm that introduces...

16. Deep Learning Algorithms on SWIR Images for Road Perception in Harsh Weather

Poster

Poster Flash Talks

Current autonomous driving algorithms heavily rely on the visible spectrum, which is prone to performance degradation in adverse conditions like fog, rain, snow, glare, and high contrast. Short-wave infrared (SWIR) has opened up new perspectives to enhance perception in such situations, which also offers several advantages over NIR and LWIR bands. This project aims to analyze the feasibility...

10. Easily Computed Marginal Likelihoods for Multivariate Mixture Models Using the THAMES Estimator

Poster

Poster Flash Talks

We present a new version of the truncated harmonic mean estimator (THAMES) for univariate or multivariate mixture models. The estimator computes the marginal likelihood from Markov chain Monte Carlo (MCMC) samples, is consistent, asymptotically normal and of finite variance. In addition, it is invariant to label switching, does not require posterior samples from hidden allocation vectors, and...

24. Exploring Bubble Flow with Deep Learning: A Preliminary Study

Poster

Poster Flash Talks

In a bubble flow study, the bubble size measurement requires identifying every instance in each image frame in the first place. However, their tininess and highly overlapped instances hindered not only human-level manipulation but also any deep learning-based detectors. In addition, their dense presentation in the observation zone made manually measuring phase time-consuming and laborious....

21. Finding missing data by modelling and solving the MCC problem

Poster

Poster Flash Talks

Missing data is a pervasive issue in empirical datasets, arising when individuals or data-collection devices fail to record observations, resulting in missing attribute values or, in some cases, entire records. Such incompleteness is common in real-world domains, including clinical databases such as Traumabase. A central challenge is whether missing values can be imputed in a manner that...

26. MMW Radar and Optical Camera Fusion Architecture for Mobile Robot Perception in Poorly Structured Environments and Adverse Weather Conditions

Poster

Poster Flash Talks

The complementary nature of radar and camera sensors offers a promising way to overcome the current limitations of robotic perception, particularly in poorly structured environments or those subject to weather disturbances. Although this synergy is recognized, research still focuses primarily on optimizing perception algorithms or decision-making systems, relegating the joint use of these...

12. Security and Confidentiality in Retrieval-Augmented Generation Systems

Poster

Poster Flash Talks

Résumé
This Ph.D. research is conducted within the LIMOS laboratory (UMR 6158), Université Clermont Auvergne, and can be part of the DATA programme (I-SITE CAP 20-25), within the domain of Artificial Intelligence for the secure exploitation of data.

The DATA programme aims to address the full lifecycle of data, from acquisition and integration to large-scale storage and value creation...

19. Vaccination des modèles de détection d'intrusion

Poster

Poster Flash Talks

Les modèles de détection d'intrusion sont sensible à l'ajout de nouveaux éléments dans leur infrastructure. On explore dans notre travail en cours la possibilité d'intégrer un échantillon d'un flux futur qu'on peux simuler dans l'infrastructure, afin d'entraîner le modèle en amont et accroître les performances. On propose ainsi un schéma de modèle semi-supervisé online de détection d'intrusion...