HAL - La Rochelle Université, Archives Ouvertes
Not a member yet
19516 research outputs found
Sort by
A small increase in temperature could rapidly expand the latitudinal range and pathogenicity of the marine trematode parasite Curtuteria arguinae
No full textInternational audienceThe phenology and transmission dynamics of the marine trematode Curtuteria arguinae were investigated, a parasite whose recently elucidated life cycle involves oystercatchers (Haematopus ostralegus), needle snails (Bittium reticulatum), and edible cockles (Cerastoderma edule). To assess the influence of temperature and light on cercarial emergence, infected B. reticulatum were exposed to temperatures ranging from 15 °C to 27 °C under both light and dark conditions. Cercarial emergence peaked under warm and illuminated conditions, showing strong temperature dependence, and, to a lesser extent, light dependence. Emergence occurred significantly at temperatures ≥21 °C, with median emergence rates of 12–47 cercariae.day−1, and the positive influence of light was overruled at higher temperatures. A 22-year dataset from Arcachon Bay (France) was also analyzed to determine the timing of first infections in cockles relative to sea and air temperatures, as well as cockle shell size, across ten cohorts. First infections of cockles were consistently observed in late summer, with a mean sea temperature of 20.8 °C, placing cercarial transmission once a thermal threshold is reached. This thermal dependency explains the parasite's current southern distribution. With climate warming, its range is likely to expand northward along the northeastern Atlantic coast, in areas where all hosts are already present. Expansion may also be supported by the northward extension of B. reticulatum. In areas where the parasite is already established, rising temperatures may enhance cercarial output and infection intensity, increasing pathogenic effects on both intermediate hosts and posing a growing threat to ecosystem functioning and cockle fisheries
LPV/Polytopic Stabilization Control and Estimation in Robotics
No full textInternational audienceNonlinear robotic systems often operate under widely varying conditions that challenge traditional linear control approaches. The Linear Parameter-Varying (LPV) paradigm overcomes these limitations and offers a unifying framework by representing the system’s time-varying dynamics as a convex blend of linear models. This enables both controller and observer synthesis through convex optimization, while considering nonlinearities and time-dependent behavior. This paper presents a linear matrix inequality (LMI)-based methodology for simultaneous stabilization control and state estimation in robotic application within the LPV/polytopic setting. Parallel to controller design, the full-state estimation challenge posed by limited sensors in robotics is addressed. An LPV observer architecture, based on the Luemberger observer, is proposed. The simultaneous observer/controller gains synthesis is then reduced to an LMI feasibility problem. The efficacy of our approach is then demonstrated and illustrated through simulations
« E-spect@tor : Enjeux de déploiement d’un outil numérique d’apprentissage individuel et collectif pour l’éducation au théâtre dans les milieux institutionnels, étude théorico pratique et approche comparatiste trans-littorale. »
No full textInternational audienceDans le cadre de l’informatisation de la recherche en Arts de la scène (Chantraine-Braillon, 2022), l’outil e-spect@tor a été développé pour former les publics à l’analyse du spectacle vivant à partir d’archives vidéo. Cette application web permet le visionnage et l’annotation chronométrée de vidéos de théâtre. Dans le cadre de notre doctorat, nous travaillons au déploiement de cet outil en milieux scolaires (écoles élémentaires et collèges) en France et en Argentine. À travers une étude théorico pratique, nous cherchons à évaluer les possibilités qu’offre e-spect@tor dans un contexte d’utilisation scolaire : quelles sont ses fonctionnalités ? Quels usages et quels objectifs avec une classe du premier ou du second degré ? Comment l’utiliser pour permettre ou faciliter l’analyse de spectacles vivants
QUEST: Quality-Aware Semi-supervised Table Extraction for Business Documents
No full textInternational audienceAutomating table extraction (TE) from business documents is critical for industrial workflows but remains challenging due to sparse annotations and error-prone multi-stage pipelines. While semi-supervised learning (SSL) can leverage unlabeled data, existing methods rely on confidence scores that poorly reflect extraction quality. We propose QUEST, a Quality-aware Semi-supervised Table extraction framework designed for Business documents. QUEST introduces a novel quality assessment model that evaluates structural and contextual features of extracted tables, trained to predict F1 scores instead of relying on confidence metrics. This quality-aware approach guides pseudo-label selection during iterative SSL training, while diversity measures (DPP, Vendi score, Int-Div) mitigate confirmation bias. Experiments on a proprietary business dataset (1k annotated + 10k unannotated documents) show QUEST improves F1 from 64% to 74% and reduces empty predictions by 45% (12% to 6.5%). On the DocILE benchmark (600 annotated + 20k unannotated documents), QUEST achieves a 50% F1 score (up from 42%) and reduces empty predictions by 19% (27% to 22%). The framework's interpretable quality assessments and robustness to annotation scarcity make it particularly suited for business documents, where structural consistency and data completeness are paramount.</div
Starting from the bottom: Contrasted trophic transfer of antibiotics and pesticides through a riverine food web
No full text(IF 8;Q1)International audienceAmong the various forms of pollution arising from human activities, the release of trace elements, pesticides and pharmaceutical residues in freshwater poses a major threat to the health and stability of these ecosystems. Once ingested by aquatic organisms, some pollutants bioaccumulate over time and biomagnify along trophic webs, while others are metabolised and/or eliminated. Understanding the trophodynamics of trace elements, pesticides and antibiotics according to their chemical properties is thus crucial. In this study, we investigated the occurrence and fate of mercury (Hg), well-studied in trophodynamic studies, pesticides and antibiotics, much less investigated despite being biologically-active compounds, in the trophic webs of the Orge River, a tributary of the Seine River, France. We sampled primary producers, macroinvertebrates, and fish, as well as sediments and water, and searched a total of 42 chemicals. We detected 19 compounds in biological samples, 9 in sediments, and 20 in water. The fate of pollutants along the trophic chain varied, with the biomagnification of Hg as predicted, and a reduction in number of antibiotics and pesticides detected along the trophic chain. This pattern suggests the metabolism and excretion along the food chain. Pollutant concentrations in biological samples were best explained by chemical properties, with LogKOW emerging as the main predictor in our models, while other tested variables (molecular weight, solubility, LogKOC, and KM) did not show significant effects. The differential transfer of pollutants along the trophic chain and the influence of chemical and biotic properties underscore the complex interactions between environmental pollution and living organisms
Forces capillaires et intégrale première de l'équation de Young-Laplace
No full textInternational audienceForces capillaires et intégrale première de l'équation de Young-Laplac
PANDORA: An Open-Access Database of Indoor Pollutant Emission Rates for IAQ Modeling
No full textInternational audienceModeling indoor air quality requires reliable data on pollutant emission rates (ERs) from indoor sources. While many studies focus on measuring indoor pollutant concentrations, far fewer provide the source-specific ERs needed for predictive modeling, and those that do often report fragmented and non-standardized formats that limit their use. This paper addresses this gap by introducing PANDORA (a comPilAtioN of inDOor aiR pollutAnt emissions), an internet-based open-access database designed to improve consistency and transparency in indoor air quality assessments. PANDORA systematically compiles ERs data for gaseous and particulate pollutants from a wide range of indoor sources. It classifies 747 sources into comprehensive categories such as construction and decoration materials (354), furniture (38), cleaning products and air fresheners (123), occupants and occupant activities (134), heating and cooking appliances (48), electrical equipment (40), whole room or building (6) and others (4). In this paper, we summarize key experimental methods used to assess the pollutants. To aid in informed decision-making, statistical analyses are provided for selected indoor pollutants of interest, including PM2.5, formaldehyde, benzene, and TVOC. Additionally, we compare the impact of using three different modeling approaches and assumptions through a case study that uses the PANDORA data to evaluate indoor pollutant ERs in a room. This application shows how PANDORA supports more transparent and consistent use of emission rate data. Our findings highlight that, despite compiling 9968 emission rate entries, expanding PANDORA with new measurements will further strengthen the accuracy and reliability of indoor air quality modeling and exposure assessments
Le café du capitaine Georgeon: horizon, liaison, héritage (1771-1832)
No full textInternational audienc
Cancer Prevalence across Vertebrates
No full textInternational audienceAbstract Cancer is pervasive across multicellular species, but what explains the differences in cancer prevalence across species? Using 16,049 necropsy records for 292 species spanning three clades of tetrapods (amphibians, sauropsids, and mammals), we found that neoplasia and malignancy prevalence increases with adult mass (contrary to Peto’s paradox) and somatic mutation rate but decreases with gestation time. The relationship between adult mass and malignancy prevalence was only apparent when we controlled for gestation time. Evolution of cancer susceptibility appears to have undergone sudden shifts followed by stabilizing selection. Outliers for neoplasia prevalence include the common porpoise (<1.3%), the Rodrigues fruit bat (<1.6%), the black-footed penguin (<0.4%), ferrets (63%), and opossums (35%). Discovering why some species have particularly high or low levels of cancer may lead to a better understanding of cancer syndromes and novel strategies for the management and prevention of cancer. Significance: Evolution has discovered mechanisms for suppressing cancer in a wide variety of species. By analyzing veterinary necropsy records, we can identify species with exceptionally high or low cancer prevalence. Discovering the mechanisms of cancer susceptibility and resistance may help improve cancer prevention and explain cancer syndromes. See related commentary by Metzger, p. 1