Publications
-
Contribution à la caractérisation de l’affordance d’un environnement de travail industriel : une approche basée sur l’apprentissage profond combinant données réelles et synthétiques
Ce travail s’inscrit dans le cadre du projet « École De La Batterie », dont l’un des objectifs concerne l’optimisation de la conception des postes de travail manuel dans le but d’améliorer leur ergonomie. Nos travaux s’inscrivent dans cette démarche et visent à caractériser l’affordance des éléments de ces environnements avec lesquels les opérateurs interagissent (outils, composants, […]
-
GZSL-MoE: Apprentissage Généralisé Zéro-Shot basé sur le Mélange d’Experts pour la Segmentation Sémantique de Nuages de Points 3D Appliqué à un Jeu de Données d’Environnement de Collaboration Humain-Robot
Résumé L’approche d’apprentissage génératif zéro-shot ou zéro exemple de données (Generative Zero-Shot Learning, GZSL) a démontré un potentiel significatif dans les tâches de la segmentation sémantique de nuages de points 3D. GZSL approche exploite des modèles génératifs comme les GAN pour synthétiser des caractéristiques réalistes (caractéristiques réelles) des classes non vues. Cela permet au modèle […]
-
From Individuals to Teams: A Conceptual Model for Group Behavior Integration in Industry 5.0
In Industry 5.0 environments, integrating human behavior models into scheduling and planning is essential. However, most research in this field focuses on individual modeling, overlooking the significant influence of group dynamics. Beyond isolated interactions, collective decisionmaking and team behaviors shape system performance, especially when deploying collaborative technologies such as Autonomous Intelligent Vehicles (AIVs) and smart […]
-
Cross-Graph Relational Knowledge Distillation with Lightweight ST-GCN for Gait Disorder Recognition
Gait recognition is essential for early diagnosis of movement disorders. The integration of new technologies can enhance the early identification of these conditions. Many current studies use Spatio-Temporal Graph Convolutional Networks (ST-GCN) that depend on skeletal data, however, these models often require substantial memory, which limits their use in clinical settings. This work introduces an […]
-
Optimizing building envelope design across various French climates: A multi-objective approach using NSGA II and RMP method
Architects face a major challenge in designing buildings that enhance human comfort while minimizing energy consumption. To address this, the present work presents a novel multi-objective optimization approach, aiming to determine the optimal building envelope design. The developed approach focuses on minimizing energy consumption for both heating and cooling demand. Therefore, the methodology follows three […]
-
nfluence de la culture organisationnelle sur la mobilité durable des étudiant.e.s
La transition vers des modes de vie plus durables ne peut se limiter à des changements individuels. Elle doit être pensée de manière systémique en intégrant l’influence de l’environnements organisationnels. Cette étude explore le lien entre la culture organisationnelle verte des établissements d’enseignement supérieur – comprenant leurs normes, valeurs et pratiques – et le choix […]
-
Life Cycle Assessment of Vernacular Construction Techniques: Comparative Analysis of Tilestone Roofing vs. Conventional Systems
Traditional tilestone roofing, commonly used in UNESCO-listed French regions such as the Causses and Ce vennes, may provide notable sustainability benefits, yet comparative studies on roofing remain scarce. Previous research has highlighted the advantages of traditional construction, especially in wall systems: As reported by [1], up to a 91% reduction in embodied emissions with natural […]
-
Impact of Successive Layer Spraying of Polymer Binder on Hygroscopic Performance of Earth Walls in Humid Tropical Environments
The use of earth in tropical building is gaining increasing attention due to its natural ability to regulate humidity, which is essential for thermal comfort. This study evaluates the hygroscopic performance of an earth wall stabilized with hemp fibers and a polymer binder, applied through successive layer spraying via gravity-driven impregnation—a technique borrowed from the […]
-
Metaheuristic and Reinforcement Learning Techniques for Solving the Vehicle Routing Problem: A Literature Review
The Vehicle Routing Problem remains a pivotal challenge in combinatorial optimization, where the objective is to determine optimal routes for a fleet of vehicles serving geographically distributed customers under specific constraints. Over decades, a diverse spectrum of solution methodologies—spanning exact algorithms, heuristics, metaheuristics, and, more recently, machine learning—has emerged. This review critically examines the intersection […]
-
A Novel Approach for Optimal Power Smoothing in Floating Offshore Wind Turbine Conversion Chains
The study proposes using battery storage systems to smooth floating offshore wind turbine (FWOT) power. However, FOWT, battery, and grid interact complexly; therefore, power flow should be optimized. This research provides a power management method that manages power flow between the FOWT and battery to smooth power grid injection.
-
An innovative Machine Learning model for predicting compressive strength of biobased concretes
Biobased concretes, which incorporate renewable and environmentally friendly components such as plant-based aggregates, offer a promising alternative to conventional materials. However, their widespread adoption is hindered by several challenges such as variability in raw materials, complex interactions between components, the lack of standardized methodologies, and requirement of advanced technics for characterizing and optimizing their mechanical […]
-
Lightweight Deep Learning for Photovoltaic Energy Prediction: Optimizing Decarbonization in Winter Houses
This paper proposes an innovative hybrid multivariate deep learning approach to predict photovoltaic (PV) energy production in winter houses, with a focus on lightweight models with low environmental impact. A methodology is developed to assess the carbon footprint of these models, considering training energy consumption, operational CO2 emissions, and energy savings from PV production optimization. […]