HAL - La Rochelle Université, Archives Ouvertes
Not a member yet
19516 research outputs found
Sort by
Anomalous Sound Detection Based on Graph Neural Networks for Forest Preservation
No full textInternational audienceAnomalous Sound Detection (ASD) has been recently extended to a range of surveillance problems where sound can be ubiquitous, such as landscape sounds, biodiversity conservation and preservation of natural environments such as forests. Tree logging was therefore given special attention, as the sounds of chainsaws can be considered anomalous in a natural environment, especially in remote and inaccessible areas. In this context, this work aims to develop an ASD model based on Graph Neural Networks (GNNs) to improve the performance of detecting anomalous sounds for illegal tree cutting, mainly produced by chainsaws. To this end, two GNN-based methods are proposed, namely a first method using Graph Convolution Networks (GCN) and an improved method using Graph Attention Networks (GAT). Experiments confirm the suitability of GNNs, particularly those based on graph attention networks, improving significantly the anomaly detection scores, with an accuracy of 0.91, an AUC score of 0.89 and high and balanced precision and recall scores, thus outperforming standard audio classification methods such as those based on Convolutional Neural Networks (CNN). Furthermore, the GAT-based model is able to rely on a minimal set of acoustic features, which paves the way for a lightweight graph model for ASD
Synthesis and Characterization of λ-Carrageenan Oligosaccharide-Based Nanoparticles: Applications in MRI and In Vivo Biodistribution Studies
No full textInternational audienceEpstein–Barr virus (EBV) infects 95% of the world’s population and persists latently in the body. It immortalizes B-cells and is associated with lymphomas. LCLs (lymphoblastoid cell lines, EBV latency III B-cells) inhibit anti-tumoral T-cell response following PD-L1 overexpression (programmed death-ligand 1 immune checkpoint). Many cancer cells, including some DLBCLs (diffuse large B-cell lymphomas), also overexpress PD-L1. Immunotherapies are based on inhibition of PD-L1/PD-1 interactions but present some dose-dependent toxicities. We aim to find new strategies to improve their efficiency by decreasing PD-L1 expression. Fucoidan, a polysaccharide extracted from brown seaweed, exhibits immunomodulatory and anti-tumor activities depending on its polymerization degree, but data are scarce on lymphoma cells or immune checkpoints. LCLs and DLBCLs cells were treated with native fucoidan (Fucus vesiculosus) or original very-low-molecular-weight fucoidan formulas (vLMW-F). We observed cell proliferation decrease and apoptosis induction increase with vLMW-F and no toxicity on normal B- and T-cells. We highlighted a decrease in transcriptional and PD-L1 surface expression, even more efficient for vLMW than native fucoidan. This can be explained by actin network alteration, suggesting lower fusion of secretory vesicles carrying PD-L1 with the plasma membrane. We propose vLMW-F as potential adjuvants to immunotherapy due to their anti-proliferative and proapoptotic effects and ability to decrease PD-L1 membrane expression
Contrat d'assurance - Indemnité d'assurance indue : droit à restitution de l'assureur sans exigence d'une autre preuve
No full textInternational audienc
Accident de la circulation - Procédure de l'offre : offre définitive présentée en cours d'instance par l'assureur
No full textInternational audienc
Investigating the quality of European silver eels by quantifying contaminants and parasite infestation in a French Mediterranean lagoon complex
No full textInternational audienceCoastal lagoons are diverse habitats with significant ecological gradients, which provide crucial ecosystem services but face threats from human activities such as invasive species and pollution. Among the species inhabiting the lagoons, the critically endangered European eel (Anguilla anguilla) is an emblematic species strongly impacted by contamination and parasitism. Several indicators were developed to assess the quality of eel at a large geographic scale. Most indicators are based on the concentration of individual pollutant and/or abundance of parasites separately without considering individual variations. This study assessed the quality of 59 eels captured at three different sites inside a Mediterranean lagoon complex (the Camargue, South of France), by integrating multiple degradation factors (POPs, TEs, and A. crassus infestation) and considering individual eel characteristics (length, age, growth rate, and sex). Using multivariate TOPSIS analysis including these degradation factors, this study found that eel quality decreased with age but did not significantly vary between sites. When focusing on each degradation factor, A. crassus infestation rates were lower in older eels, independently to the site, however, the POPs and TEs contaminations were lower in the Grandes Cabanes site compared to the Vaccarès and Fumemorte sites even if larger and younger eels were more contaminated by POPs. These findings reveal the finescale spatial variability in eel quality, with TOPSIS analysis providing a robust method to rank and score scenarios. This approach enhances the understanding of habitat degradation sources affecting eel contamination and parasitic infestation, supporting more effective strategies for sustainable habitat management
Missing data estimation method for durability survey of reinforced concrete structures
No full textInternational audienceReinforced concrete structures are well-known for their high durability, however, they remain vulnerable to natural hazards and extreme events that can impact their performance over time. In aggressive environments, there is a high likelihood of increased maintenance, rehabilitation, and repair actions that constitute a significant portion of the total lifecycle spending. Monitoring systems have been implemented during the last decades to collect periodically or continuously essential data about the durability performance of the structures in real operation. However, the effectiveness of these systems is impacted by sensor efficacy, influenced in turn by environmental factors, sensor durability, and power outages, leading to intermittent or permanent data gaps. This study proposes a methodology to address the problem of missing data of a Structural Health Monitoring (SHM) system, specifically aiming to provide more accurate and continuous information from concrete resistivity and temperature sensors to support the early detection of corrosion. The proposed methodology was applied to a repaired reinforced concrete structure with over fourteen years of data, where significant gaps in the measurements were present. The approach combines several techniques to fill these gaps: deep machine learning for air temperature, generalized linear models for concrete temperature, and pattern recognition for concrete resistivity. To the best of the authors' knowledge, this is the first time a methodology has been proposed for imputing missing data from resistivity sensors in SHM systems, which are increasingly being implemented. This approach is innovative and offers potential benefits for SHM system managers, providing more information on long-term sensor data that could aid in early corrosion detection and maintenance planning. The application of the proposed methodology to a real case study indicated a successful imputation of 43.4% of missing data although some challenges persist for sensors located in areas characterized by high measurements variability
La sorisete des estopes - Édition, traduction et notes d'après le manuscrit Bern, Burgerbibliothek, Cod. 354
No full textÉdition, traduction et notes de La sorisete des estopes d'après le ms. Bern, Burgerbibliothek, Cod. 35
"Le sentier batu" de Jean de Condé - Édition, traduction et notes d'après le manuscrit Arsenal 3524
No full textÉdition, traduction et notes du Sentier batu de Jean de Condé d'après le ms. Arsenal 352
Le moigne - Édition, traduction et notes d'après le manuscrit BnF Rothschild 2800
No full textÉdition, traduction et notes du Moigne d'après le manuscrit BnF Rothschild 280