HAL-UNILIM
Not a member yet
39506 research outputs found
Sort by
La réforme de l’action de groupe, ou l’art de « couper la poire en deux »
No full textInternational audienceLa réforme attendue de l’action de groupe est entrée en vigueur le 3 mai 2025. Profitant de la transposition de la directive (UE) n° 2020/1828 relative aux actions représentatives, le législateur a opéré une large refonte de la directive (UE) n° 2020/1828 relative aux actions représentatives, le législateur a opéré une large refonte de ce mécanisme procédural singulier. Si la nouvelle loi apporte des améliorations notables par rapport au de ce mécanisme procédural singulier. Si la nouvelle loi apporte des améliorations notables par rapport au droit antérieur, elle n’en demeure pas moins imparfaite sur certains aspects essentiels à l’efficacité du droit antérieur, elle n’en demeure pas moins imparfaite sur certains aspects essentiels à l’efficacité du dispositif
Antibiotic susceptibility according to age of clinical strains from hospital respiratory samples: A nationwide study
No full textInternational audienceObjectives: This study evaluated the microbiological distribution and sensitivity of clinical strains from respiratory samples of inpatients to the antibiotics usually recommended for treatment of lower respiratory tract infection (LRTI).Methods: Using the French SPARES (Surveillance et Prévention de l'AntibioRésistance en Etablissement de Santé) database, we recorded all respiratory microbiological samples collected in 2022 in 409 hospitals located in all regions of France. The distribution of main bacterial species and their resistance to the most frequently prescribed antibiotics for LRTI were compared by age group (18-64 years, 65-79 years, ≥ 80 years).Results: Among 48,721 strains, 47.7 % were isolated from patients aged 18-64 years, 37.9 % aged 65-79 years, and 14.4 % aged ≥ 80 years. Enterobacteriaceae (30.1 %, 36.0 % and 35.1 %, respectively) and Pseudomonas aeruginosa (19.0 %, 25.0 % and 27.1 %) were the most prevalent pathogens, especially in older patients. Conversely, Haemophilus influenzae (14.5 %, 10.7 % and 8.7 %) and Streptococcus pneumoniae (8.3 %, 5.6 % and 3.4 %) were rare in older age. Overall antibiotic resistance increased with age across all classes: in increasing order of resistance, levofloxacin (9.2 %, 11.1 % and 13.2 %), piperacillin-tazobactam (14.5 %, 16.8 % and 17.4 %), cefotaxime (27.4 %, 34.3 % and 39.5 %), doxycycline (28.2 %, 37.4 % and 39.5 %), cotrimoxazole (32.1 %, 38.5 % and 40.2 %), and amoxicillin-clavulanate (37.2 %, 46.0 % and 51.4 %, p < 0.05 for all comparisons). More than half of the strains were resistant to amoxicillin and erythromycin.Conclusions: In this large nationwide database of respiratory samples, older age was associated with a high prevalence of Enterobacteriaceae and P. aeruginosa and beta-lactam resistance, a finding challenging current LRTI probabilistic treatment. Conversely, H. influenzae and S. pneumoniae were rarely observed in patients over 80 years of age
Distributed alternating direction method of multipliers approach for the power distribution network reconfiguration
No full textInternational audienceThe electrical network reconfiguration problem aims to minimize losses in a distribution system by adjusting switches while ensuring radial topology. The growing use of renewable energy and the complexity of managing modern power grids make solving the reconfiguration problem crucial. Distributed algorithms help optimize grid configurations, ensuring efficient adaptation to changing conditions and better utilization of renewable energy sources. This paper introduces a distributed algorithm designed to tackle the problem of power distribution network reconfiguration with a radiality constraint. This algorithm relies on ADMM (Alternating Direction Method of Multipliers), where each agent progressively updates its estimation based on the information exchanged with neighboring agents. We show that every agent is required to solve a linearly constrained convex quadratic programming problem and a Minimum Weight Rooted Arborescence Problem (MWRAP) with local weights during each iteration. Through numerical experiments, we demonstrate the performance of the proposed algorithm in various scenarios, including its application to a 33-bus test system and a real-world network
Clock mechanism generation using interaction graphs
No full textInternational audienceThe 3D modeling of heritage objects is now regarded as a significant aid and has become essential in the study and mediation of cultural heritage. In horology, 3D models serve not only for visualization but also to simulate complex movements, replicate mechanisms, and study their functionality for scientific and educational purposes. While promising, modeling accurate functional 3D clock models presents challenges that require considerable time and expertise in both clockmaking and mechanical modeling. This paper presents a new method to automatically generate functional mechanical clock assemblies given a formalized interaction graph. The latter describes both the parts of the mechanism and the sub-assemblies, all respecting constraints defined with the help of experts in the field. This graph is then used as a base for our novel parameter propagation algorithm to automatically generate valid 3D models of the corresponding clock from a limited amount of input data. Our method provides different generations of clocks that adhere to functional constraints such as the mechanism category (e.g. comtoise, Neuchâteloise, Paris...), gear ratios or part placement
Artificial intelligence in personalized rehabilitation: current applications and a SWOT analysis
No full textInternational audienceArtificial intelligence (AI) is transforming personalized rehabilitation by introducing innovative methods to enhance care across diverse medical specialties. Despite its potential, widespread implementation remains limited, largely due to a lack of comprehensive analyses on its benefits and barriers. This mini narrative review examines current applications of AI in personalized rehabilitation and provide a SWOT (Strengths, Weaknesses, Opportunities, Threats) analysis AI is already being used to develop personalized treatment plans, support ongoing patient management, and adapt therapy sessions in real-time. One of its key strengths is the capacity to process vast datasets and monitor real-time information, thereby elevating the level of personalization. Automation of certain tasks can reduce human error and alleviate clinician workload, allowing more time for direct patient care. Opportunities for AI lie in leveraging rapidly advancing technologies to meet the rising demand for rehabilitation services, particularly with aging populations. Collaborations with industry can accelerate innovation, while data sharing can promote best practices across institutions. However, notable challenges persist. High implementation costs, ethical concerns such as algorithmic bias, and risks of increasing healthcare disparities remain major barriers. Additionally, threats such as data privacy breaches and security vulnerabilities emphasize the need for robust, balanced regulatory frameworks. In conclusion, AI holds immense promise for transforming personalized rehabilitation. While current applications are largely in early stages or proof-of-concept phases, ongoing research, ethical foresight, and strategic collaboration are essential to maximize benefits and minimize risks for optimal patient outcomes
Features of Invasive Aspergillosis Caused by Aspergillus flavus , France, 2012–2018
No full textInternational audienceInvasive aspergillosis (IA) caused by Aspergillus flavus remains poorly described. We retrospectively analyzed 54 cases of IA caused by A. flavus reported in France during 2012–2018. Among cases, underlying IA risk factors were malignancy, solid organ transplantation, and diabetes. Most (87%, 47/54) infections were localized, of which 33 were pleuropulmonary and 13 were ear-nose-throat (ENT) infection sites. Malignancy (70% [23/33]) and solid organ transplantation (21% [7/33]) were the main risk factors in localized pulmonary infections, and diabetes mellitus was associated with localized ENT involvement (61.5%, [8/13]). Fungal co-infections were frequent in pulmonary (36%, 12/33) but not ENT IA (0 cases). Antifungal monotherapy was prescribed in 45/50 (90%) cases, mainly voriconazole (67%, 30/45). All-cause 30-day case-fatality rates were 39.2% and 90-day rates were 47.1%, and rates varied according to risk factor, IA site, and fungal co-infections. Clinicians should remain vigilant for A. flavus and consider it in the differential diagnosis for IA
Embedding similarity guided license plate super resolution
No full textInternational audienceSuper-resolution (SR) techniques play a pivotal role in enhancing the quality of low-resolution images, particularly for applications such as security and surveillance, where accurate license plate recognition is crucial. This study proposes a novel framework that combines pixel-based loss with embedding similarity learning to address the unique challenges of license plate super-resolution (LPSR). The introduced pixel and embedding consistency loss (PECL) integrates a Siamese network and applies contrastive loss to force embedding similarities to improve perceptual and structural fidelity. By effectively balancing pixel-wise accuracy with embedding-level consistency, the framework achieves superior alignment of fine-grained features between high-resolution (HR) and super-resolved (SR) license plates. Extensive experiments on the CCPD and PKU dataset validate the efficacy of the proposed framework, demonstrating consistent improvements over state-of-the-art methods in terms of PSNR, SSIM, LPIPS, and optical character recognition (OCR) accuracy. These results highlight the potential of embedding similarity learning to advance both perceptual quality and task-specific performance in extreme super-resolution scenarios
Modulation of the RE3NbO7 composition to stabilize cubic varieties under a high temperature treatment: structural and spectroscopic characterizations
No full textInternational audienceIn the framework of the development of new optical compounds..
Extreme ultraviolet source based on high harmonic generation in solids with a high-energy fiber-based 1550 nm driving laser
No full textNational audienceHigh harmonic generation (HHG) is a well-established process used to produce light in the extreme ultraviolet (EUV) range. EUV sources hold significant potential for applications in ultrafast spectroscopy, lithography, and atom probe tomography. High harmonics of the driving laser frequency can be generated by tightly focusing a high-energy laser into a nonlinear medium, such as a gas, solid, or plasma. Historically, HHG has been primarily driven by lasers with wavelengths around 800 nm. However, mid-infrared (mid- IR) lasers have gained increasing interest due to their ability to extend the harmonic cutoff, enabled by the quadratic wavelength scaling of the ponderomotive energy [1]. Additionally, mid-IR drivers offer improved damage thresholds, making them attractive for high-intensity applications. Optical Parametric Chirped-Pulse Amplification (OPCPA) and Optical Parametric Amplifier (OPA) systems have proven effective in achieving these results, but their inherent complexity highlights the need for more compact, efficient, and scalable EUV generation techniques [1, 2, 3]. In this context, ultrafast mid- IR fiber lasers have emerged as a promising alternative for HHG. However, their pulse energies remain limited to the nanojoule range, posing a challenge for generating high-flux EUV radiation [4]. In this work, we demonstrate high-order harmonic generation in bulk crystals using a high-energy fiber laser system operating near 1550 nm. The system delivers few-cycle pulses with durations below 50 fs, microjoule- level pulse energies, and a repetition rate of 660 kHz. This is achieved by combining a high-energy chirped- pulse amplifier featuring erbium-doped fibers, initially producing 600 fs pulses, with a post-compression stage based on an Argon-filled hollow-core photonic crystal fiber (HC-PCF) [5, 6]. The interplay between anomalous dispersion and Kerr nonlinearity along the HC-PCF enables the generation of fundamental solitons with more than 18 Megawatts peak power and sub-50 fs duration. The compressed pulses are then focused into an MgO crystal to generate EUV high-order harmonics. The EUV spectra were measured using a home-built spectrometer and analyzed as a function of the crystal orientation relative to the laser polarization, and the driving laser intensity. As a result, harmonic orders up to the 31st were detected, with a total photon flux reaching 5 × 10⁶ photons per second. To conclude, we demonstrated the successful use of the high-energy fiber laser platform with a HC-PCF- based post compression stage as a driving laser for efficient HHG-based EUV source
Place accordée aux handicaps dans le narratif des clubs professionnels de basket-ball français.: Analyse des projets-clubs dans le cadre du label LNB (2013-2023)
No full textInternational audienceA Antarès, salle du club professionnel de basket masculin du Mans (Sarthe), l’image est habituelle : les personnes en situation de mobilité réduite s’installent au premier rang, entre le terrain et les premières tribunes, pour soutenir leur équipe. En ligne, la rubrique « billetterie » explique que les personnes en fauteuil peuvent accéder au match gratuitement (avec un accompagnateur bénéficiant du demi-tarif, dans la limite des places disponibles). A l’adidas arena, où évolue le Paris Basket, les places en bord de terrain, nominatives, sont appelées « courtside seats » et intégrées à l’offre d’hospitalités. Le champagne y est fourni à discrétion. Le site internet du club propose de fournir les tarifs à la demande …Entre ces deux situations, la question se pose de la place accordée aux handicaps par les clubs professionnels de basket français. Cette recherche se propose d’étudier les projets-clubs rédigés dans le cadre du processus de demande de labellisation de la ligue nationale de basket (LNB) pour tenter d’y répondre. De 2013 à 2023, les clubs volontaires de première et seconde division professionnelle ont transmis des narratifs évoquant leur stratégie de billetterie et d’exploitation des installations sportives mises à leur disposition mais également leurs actions sociétales et territoriales. L’étude longitudinale de ces narratifs sera mise en perspective avec les apports de la théorie néo-institutionnelle. Il s’agira d’une part d’observer la stabilité des organisations et leur conformité aux attentes sociales (isomorphisme institutionnel) et d’autre part de repérer les traces du changement des pressions institutionnelles (Daudigeos, 2019). Les résultats devraient permettre de compléter les travaux d’une part sur l’accessibilité des personnes en situation de handicap (Bouttet, 2019) et d’autre part sur les publics du basket professionnel (Charlot & Clément,2004)