Scientific Publications of the University of Toulouse II Le Mirail
Not a member yet
92205 research outputs found
Sort by
Sapin, hêtre, chêne en Occitanie : be or has been?
No full textInternational audienceThe ongoing climate change, in conjunction with increasing disturbances such as insect infestations, droughts, and fires, is prompting forest managers to consider the future viability of certain tree species and the most suitable management strategies for their forests. This prompts a critical question: does a given species possess the capacity to adapt to novel environmental conditions, or will it need to range into more favorable habitats to ensure its survival?In the Occitanie region, the main consequences of climate change have been an increase in average temperatures since the mid-20th century, particularly marked in spring and summer, and an increase in the frequency and intensity of droughts. The decrease in the amount and duration of snow cover is also a notable change, directly affecting water availability. Furthermore, the eastern part of the region is particularly affected by the decline in annual precipitation. The observed decline in fir, beech, and oak trees has been attributed to these climatic changes. However, these three tree species, which represent major economic resources for the regional forestry industry, now pose crucial adaptation challenges for managers, owners, and communities.Tree growth rings offer insights into the response of trees to climate change and their ability to withstand extreme conditions. As part of several research projects conducted since 2016, we have compiled a vast set of chronologies for these three tree species across Occitanie, thereby filling a significant data gap for this region. The database currently includes 14 fir populations (1605-2023), 4 beech populations (1840-2021), and 10 oak populations (1754-2024). In this presentation, we will present the initial results concerning the response of trees to climate along the different climatic gradients that characterize the region. We will also examine their resilience to the severe droughts that have occurred in recent decades. We will discuss these results in light of the global warming expected in the coming decades and how to take this into account in forest management.Les changements climatiques actuels, associés à une intensification des régimes de perturbation (attaques d’insectes, sécheresses, incendies, etc.), soulèvent de sérieuses interrogations chez les gestionnaires forestiers quant à l’avenir de certaines espèces d’arbres, mais aussi sur les stratégies de gestion – ou de non-gestion – à mettre en œuvre dans leurs forêts. Une question centrale en découle : une espèce donnée est-elle en mesure de s’adapter aux nouvelles conditions environnementales ou devra-t-elle déplacer son aire de distribution pour survivre dans des habitats plus favorables ?En région Occitanie, les principales conséquences du changement climatique se traduisent par une hausse des températures moyennes depuis le milieu du XXe siècle, particulièrement marquée au printemps et en été, ainsi qu’une augmentation de la fréquence et de l’intensité des épisodes de sécheresse. La diminution de la quantité et de la durée de l’enneigement constitue également un changement notable, affectant directement la disponibilité en eau. Par ailleurs, la partie orientale de la région est particulièrement touchée par la baisse des précipitations annuelles. Les dépérissements observés chez le sapin, le hêtre et le chêne ont été largement attribués à ces évolutions climatiques. Or, ces trois espèces d’arbres, qui représentent des ressources économiques majeures pour la filière forestière régionale, posent aujourd’hui des enjeux cruciaux d’adaptation pour les gestionnaires, propriétaires et collectivités.Les cernes de croissance des arbres fournissent des informations précieuses sur la sensibilité des arbres au climat et sur leur capacité de résilience face aux événements extrêmes. Dans le cadre de plusieurs projets de recherche menés depuis 2016, nous avons constitué un vaste jeu de chronologies pour ces trois espèces d’arbres à travers l’Occitanie, comblant ainsi un déficit majeur de données pour cette région. Cette base de données est actuellement composée de 14 populations de sapin (1605-2023), de 4 populations de hêtre (1840-2021) et de 10 populations de chênes (1754-2024). À travers cette communication, nous présenterons les premiers résultats concernant la réponse des arbres au climat le long des différents gradients climatiques qui caractérisent la région. Nous explorerons également leur capacité de résilience face aux épisodes de sécheresse sévères survenus au cours des dernières décennies. Nous discuterons ces résultats au regard du réchauffement global attendu dans les décennies à venir et de comment en tenir compte dans la gestion forestière
Nxcap: A Unified Format for NIDS Benchmarking
No full textInternational audienceThis paper introduces Nxcap, a unified network traffic floworiented capture format designed to bridge the gap between descriptive statistics and packet capture files (Pcap) for network intrusion detection systems (NIDS) benchmarking. Nxcap represents network traffic as a collection of flows with associated packets, using a minimal set of features that enable both Pcap reconstruction and the extraction of flow-level descriptive statistics. This format addresses the limitations of existing datasets, which often suffer from inconsistencies, lack of standardization, and errors due to misconfigurations or biases introduced during data collection. We present libnxcap, an open-source library that captures network traffic in the Nxcap format, facilitates flow labeling, and includes mechanisms to verify and correct data at the time of capture. A key contribution is the advanced deduplication engine, which operates at the flow level and accounts for protocol behavior, revealing significant duplication issues in popular datasets like UNSW-NB15. Experiments demonstrate that Nxcap reduces dataset storage requirements by over 98% compared to raw Pcap files, while ensuring data quality and compatibility with existing tools. This work aims to provide the research community with a standardized, efficient, and reliable format for network traffic representation, enhancing the development and evaluation of NIDS
On the instability of local learning algorithms: Q-learning can fail in infinite state spaces
No full textWe investigate the challenges of applying model-free reinforcement learning algorithms, like online Q-learning, to infinite state space Markov Decision Processes (MDPs). We first introduce the notion of Local Learning Processes (LLPs), where agents make decisions based solely on local information, and we show that Q-learning can be seen as a specific instance of an LLP. Using renewal techniques, we analyze LLPs and demonstrate their instability under certain drift and initial conditions, revealing fundamental limitations in infinite state spaces. In particular, we show that while asymptotically optimal in finite settings, Q-learning can face instability and strict sub-optimality in infinite spaces. Our findings are illustrated through queueing system examples drawn from load balancing and server allocation. The study underscores the need for new theoretical frameworks and suggests future research into nonlocal Q-learning variants
A New Geometric Regression with Inputs-Outputs on Matrix Lie Groups
No full textInternational audienceThis paper investigates a new Lie group regression model for input-output data belonging to Lie groups. The originality of the model lies in the fact that the unknown weights also lie in Lie groups and are learned using an intrinsic optimization algorithm based on maximum likelihood estimation. The model is validated through numerical simulations conducted using synthetic data belonging to the Lie group SO(3), which is commonly used in robotics to represent rotational observations
Censeur/Rabat-joie
No full textInternational audienceIn the context of the reactionary offensive on Western democracies that has taken place for about ten years, the article intends to differentiate moral and social censorship from political criticism, through the distinction between censor and killjoy – a concept theorized by feminist researcher and activist Sara Ahmed, that we rely on in the present paper. In the light of this necessary distinction, we intend to shed light on the fallacious nature of the accusation of censorship or canceling culture addressed by reactionary ideologists to minority or subordinate social groups in struggle, and to reassess the political legitimacy of debates which stir up trouble in society and university campuses about works or cultural production considered as problematic.Dans le contexte d’offensive réactionnaire qui touche les démocraties occidentales depuis une dizaine d’années, l’article entend différencier la censure morale et sociale de la critique politique, à travers la distinction entre censeur et rabat-joie – un concept théorisé par la chercheuse et militante féministe Sara Ahmed que nous reprenons à notre compte. À l’aune de cette distinction, nous entendons démontrer le caractère fallacieux de l’accusation de « censure » ou de « canceling culture » adressée par les idéologues réactionnaires aux groupes sociaux minoritaires ou subalternes en lutte, et réévaluer la légitimité politique des débats qui agitent la société et les campus universitaires au sujet d’œuvres ou de productions culturelles considérées comme problématiques
L’écologie politique de l’huître.: Ménager les ressources sous-marines de la rade de Brest au milieu du XIXe siècle
No full textInternational audienceThe flat oyster ( Ostrea edulis ), today being rediscovered for its properties as an ecosystem engineer and sentinel species, had almost disappeared from European coasts at the turn of the nineteenth and twentieth centuries, mainly due to overfishing. However, like forests, oyster beds have long been subject to regulatory measures aimed at “managing” in order to “conserve” them. Based on the study of a dispute over the use of underwater resources harvested from the seabed of Brest Roadstead, this article offers a conflictual history of conservation. Paying close attention to the actors and the archives they produced, it reconstructs the knowledge, instruments, and institutions developed by the members of a joint commission appointed in 1847 to establish a new system of exploitation within this territory located between land and sea. In so doing, it highlights the ambivalences inherent in attempts to manage nature: the concept of ménagement , omnipresent in the documentation, alternately means “using economically” (or even “taking care” of) maritime resources, “reconciling” the different interests involved in their exploitation, and, when the prefix a- is added to form aménager , “regulating” or “ordering” this exploitation in order to make it more “rational.”Aujourd’hui redécouverte pour ses propriétés d’espèce ingénieure et sentinelle, l’huître plate ( Ostrea edulis ) a presque disparu des côtes européennes au tournant du xx e siècle, principalement en raison de la surpêche. Pourtant, comme les forêts, les bancs d’huîtres ont fait de longue date l’objet de mesures de régulation qui visaient à les « ménager » afin de les « conserver ». À partir de l’étude d’un conflit d’usage autour de différentes ressources sous-marines tirées du fond de la rade de Brest, cet article propose une histoire conflictuelle de la conservation. Écrit au plus près des acteurs et des archives qu’ils ont produites, il restitue les savoirs, les instruments et les institutions élaborés par les membres d’une commission mixte d’enquête nommé en 1847 pour instaurer un nouvel ordre d’exploitation à l’échelle de ce territoire situé entre terre et mer. Ce faisant, il met en lumière les ambivalences inhérentes aux entreprises de « ménagement » de la nature, un concept omniprésent dans la documentation qui signifie tour à tour « user avec économie » des choses de la mer (voire en « prendre soin »), « concilier » les différents intérêts engagés par leur exploitation et, lorsqu’on lui ajoute le préfixe « a- » pour former « aménager », « régler » ou « mettre en ordre » cette exploitation de manière à la rendre plus « rationnelle »
Prevalence of pediatric neuromuscular disorders in the Southwest region of France
No full textInternational audienceAround the world, epidemiological data on neuromuscular disorders (NMD) in the pediatric population are very limited. Through medical and genetic hospital records in expert NMD centers (Toulouse, Montpellier, Bordeaux), from May 2001 to June 2022, NMD pediatric prevalence and the epidemiological profile of these disorders were investigated. We performed a retrospective cohort study with data from the French National Rare Disease Databank, which gathers a minimal dataset on all patients. The prevalence by diagnosis and age group or by year and survival from birth for muscular disorders by sub-group were analyzed. 1621 children were included with 62 % males. We estimated the regional prevalence at 37.9 (CI95 % = 35.3 − 40.7)/100,000 inhabitants under 18 years old. For the muscular disorder sub-cohort analysis, we estimated regional prevalence for Duchenne, Becker, Charcot-Marie-Tooth type 1 and Spinal Muscular Atrophy at 5 (CI95 % = 4.1 − 6.1), 1.3 (CI95 % = 0.9 − 1.9), 6.2 (CI95 % = 0.1 − 7.3) and 3.2 (CI95 % = 2.5 − 4.1), respectively. Our findings seem in accordance with other previous but scarce data. Together, these data may reflect a consensus across countries. This epidemiological study is the first robust estimation of the French pediatric prevalence of neuromuscular disorders and presents a strong starting point to be confirmed by analyses extended to all French expert centers
Biomechanically consistent real-time action recognition for human-robot interaction
No full textInternational audienceThis paper presents a novel framework for real-time human action recognition in industrial contexts, using standard 2D cameras. We introduce a complete pipeline for robust and real-time estimation of human joint kinematics, input to a temporally smoothed Transformer-based network, for action recognition. We rely on a new dataset including 11 subjects performing various actions, to evaluate our approach. Unlike most of the literature that relies on joint center positions (JCP) and is offline, ours uses biomechanical prior, eg. joint angles, for fast and robust real-time recognition. Besides, joint angles make the proposed method agnostic to sensor and subject poses as well as to anthropometric differences, and ensure robustness across environments and subjects. Our proposed learning model outperforms the best baseline model, running also in real-time, along various metrics. It achieves 88% accuracy and shows great generalization ability, for subjects not facing the cameras. Finally, we demonstrate the robustness and usefulness of our technique, through an online interaction experiment, with a simulated robot controlled in real-time via the recognized actions
Rolling horizon data driven robust optimization for supply chain planning
No full textInternational audienceThis paper discusses the challenge of production planning in a dyadic supply chain, where uncertainties disrupt plans that are updated through a rolling horizon DRP process. These uncertainties, such as demand fluctuations, machine failures, and delivery delays, cause instability and worsen the bullwhip effect, which reduces the reliability and effectiveness of production plans.To address these challenges, we propose a data-driven robust optimization framework that uses historical data analysis and clustering techniques to create well-defined uncertainty sets. Numerical experiments, using simulated historical data, show that this approach improves supply chain planning by balancing precision and robustness in managing uncertainties
