HAL-INSA Toulouse
Not a member yet
34325 research outputs found
Sort by
Using a pre-trained machine learning model to estimate the 3d-ground reaction forces during rugby scrummaging with instrumented insoles
No full textInternational audienceINTRODUCTION:Rugby scrummaging represents a crucial phase of the game, characterized by high-intensity physical efforts and a significant impact on match outcomes [1]. The horizontal force generated by the entire pack is a key determinant of scrum success. However, existing measurement systems are unable to provide 3D, individual, and on-field assessments of ground reaction forces (GRF). A previous study developed a Machine Learning (ML) model to predict the 3D-GRF with instrumented insoles during scrummaging but this was conducted on recreationally active subjects without specific scrummaging experience [2]. Thus, this study aimed to investigate to what extent this model can be used to predict the 3D-GRF for elite rugby players.METHODS:Twelve elite rugby players (12 males; age: 20+/-1 ans; height: 191+/-7 cm; weight: 116+/-13 kg) performed three pushing trials of 15 seconds, against a fixed scrum machine. They wore commercial instrumented insoles (Loadsol Pro®, Novel, Germany, 200Hz) inside their shoes with each foot on a force plate (Sensix, 1000Hz) covered with artificial turf. The force plate data served as the reference for 3D-GRF measurements. For each subject, one trial was used to infer data from the pre-trained ML model, while the remaining two trials were utilized to personalize the model for each subject. The model’s performance was assessed by computing the Root Mean Square Error (RMSE) between the prediction and the reference, the correlation coefficient (r), and the percentage of RMSE compared to the mean resultant force.RESULTS:The initial model inference yielded mean RMSE values of 42±9N on the Medio-Lateral (ML) axis, 168±87N on the Antero-Posterior (AP) axis, and 180±49N on the Vertical (V) axis, with correlation coefficients r of 0.708±0.110 (ML), 0.825±0.102 (AP), and 0.571±0.123 (V). RMSE percentages relative to the mean resultant force were 4.2±0.7% (ML), 15.8±5.0% (AP), and 18.0±4.8% (V). After personalization, RMSE were 32±10N (ML), 99±41N (AP), and 135±38N (V), and correlation coefficients r were0.838±0.062 (ML), 0.911±0.044 (AP), and 0.680±0.168 (V). The mean percentages of RMSE compared to the average resultant force were 3.1±0.6% (ML), 9.7±2.8% (AP), and 14.0±5.5% (V).CONCLUSION:The findings of this study indicate that an ML model pre-trained on data from recreationally active individuals without specific rugby scrummaging experience is not optimal for accurately estimating 3D-GRF in elite rugby players. Model personalization for each participant improved performance, suggesting that personalization is a promising approach for enhancing ML model performance when the model is trained on a non-specific dataset. However, further improvement in model performance could be achieved by pre-training the model on a dataset more closely aligned with the data used for personalization, to ensure reliable predictions.References[1] Scott et al., J. Sci. Med. Sport, 2023[2] Pomarat et al., IEEE Xplore, 202
LANGAGE DE SPÉCIFICATION PARALLEL-DEVS POUR LA MODÉLISATION : APPROCHE MATHÉMATIQUE ET GRAMMAIRE
No full textInternational audienceIn this article, we propose a new formal language called the Formal Parallel-DEVS Modeling Language (FPDEVSML) as a platform-independent specification of the Parallel-DEVS (PDEVS) formalism. The DEVS (Discrete Event Systems Specification) formalism enables the specification of discrete event models in a hierarchical and modular manner, providing a solid foundation for the modeling, simulation, and analysis of discrete systems. FPDEVSML is based on a grammar with a rigorous mathematical structure that formalizes the sets and mathematical objects used in PDEVS modeling. The main objective of the proposed approach is to provide a formal framework for specifying and analyzing a system and to improve interoperability between PDEVS simulators. FPDEVSML can be used as an intermediate target language for DSLs. The specified models can then be translated into different forms of code using code generators and then executed with various tools for model verification and execution.Dans cet article, nous proposons un nouveau langage formel, le Formal Parallel-DEVS Modeling Language (FPDEVSML), comme langage de spécification indépendante de la plateforme du formalisme Parallel-DEVS (PDEVS). Le formalisme DEVS (Discrete Event Systems Specification) permet de spécifier des modèles à événements discrets de manière hiérarchique et modulaire, offrant ainsi une base solide pour la modélisation, la simulation et l'analyse de systèmes discrets. FPDEVSML repose sur une grammaire dotée d'une structure mathématique rigoureuse qui formalise les ensembles et objets mathématiques utilisés dans la modélisation PDEVS. L'objectif principal de l'approche proposée est de fournir un cadre formel pour la spécification et l'analyse d'un système et d'améliorer l'interopérabilité entre les simulateurs PDEVS. FPDEVSML peut être utilisé comme langage cible intermédiaire pour les DSL. Les modèles spécifiés peuvent ensuite être traduits en différentes code des langages de programmation à l'aide de générateurs de code, puis exécutés avec divers outils de vérification et d'exécution de modèles
Synergistic C–H bond activation across molybdenum–iridium multiply bonded complexes: a cascade of transformations
No full textInternational audienceeterobimetallic compounds offer unique opportunities for activating substrates through cooperative interactions between two distinct metal centers, potentially leading to catalytic reactivity beyond the reach of monometallic systems. The pairing of molybdenum with iridium has recently shown remarkable performance in heterogeneous catalytic and electrocatalytic processes, yet remains surprisingly unexplored in the context of homogeneous catalysis and well-defined molecular complexes. In this work, we address this gap by reporting the synthesis and characterization of complexes featuring rare molybdenum–iridium multiple bonds: Cp*Ir(H)Mo(NMe2)3, 1 and (Cp*IrH2)2Mo(NMe2)2, 2. Both complexes undergo electrophilic insertions of heteroallenes (CO2, tBuNCO) into the ancillary amido ligands. In the case of compound 1, the insertion of tert-butyl isocyanate initiates a unique cascade of bond breaking/forming events – including triple C–H activations and Csp3–Csp3 coupling – ultimately yielding a Mo(VI) metallacyclopropane complex, 5. Addition of multiple equivalents of isocyanate instead interrupts this mechanism, leading to C–N and C–O bonds cleavage and the formation of aminocarbyne and imido ligands bridging the Ir and Mo centers, along with a molybdenum(VI)-oxo moiety. This unprecedented reactivity, mediated by a key reaction intermediate featuring an unsupported Mo–Ir quadruple bond, illustrates the interest of heterobimetallic compounds for complex bond activation and reorganization
Toward a Real-Time Framework for Accurate Monocular 3D Human Pose Estimation with Geometric Priors
No full textInternational audienceMonocular 3D human pose estimation remains a challenging and ill-posed problem, particularly in real-time settings and unconstrained environments. While direct imageto-3D approaches require large annotated datasets and heavy models, 2D-to-3D lifting offers a more lightweight and flexible alternative-especially when enhanced with prior knowledge. In this work, we propose a framework that combines real-time 2D keypoint detection with geometry-aware 2D-to-3D lifting, explicitly leveraging known camera intrinsics and subject-specific anatomical priors. Our approach builds on recent advances in self-calibration and biomechanically-constrained inverse kinematics to generate large-scale, plausible 2D-3D training pairs from MoCap and synthetic datasets. We discuss how these ingredients can enable fast, personalized, and accurate 3D pose estimation from monocular images without requiring specialized hardware. This proposal aims to foster discussion on bridging data-driven learning and model-based priors to improve accuracy, interpretability, and deployability of 3D human motion capture on edge devices in the wild
IsoDesign: software for optimizing the isotopic composition of labeled substrates in 13C-fluxomics experiments
No full textInternational audienceThe study of metabolic fluxes provides a detailed phenotypic and functional description of cellular metabolism, contributing to a better understanding of biological processes. Metabolic Flux Analysis (MFA) encompasses a set of approaches used to quantify biochemical reaction rates within metabolic networks, relying on mathematical modeling methods.13C-MFA is an approach that leverages isotopic data from stable carbon-13 (13C) labeling experiments to calculate metabolic fluxes. The ability to compute fluxes of interest and their accuracy strongly depend on the experimental design, particularly on the choice of labeled substrates.However, determining and selecting the optimal labeling configurations of substrates is a complex and time-consuming task, requiring the exploration of multiple labeling configurations within a vast solution space. This selection is even more challenging as it depends on optimizing flux accuracy, analytical capabilities, and the often high cost of labeled substrates, making the analysis of possible combinations even more complex.To accelerate this process, we have developed IsoDesign, an open-source Python tool (https://github.com/MetaboHUMetaToul-FluxoMet/IsoDesign) dedicated to optimizing the isotopic composition of labeled substrates for 13C-fluxomics experiments. With its intuitive graphical interface and the use of the influx si software (https://github.com/sgsokol/influx) for flux calculation, IsoDesign facilitates the design of isotopic labeling experiments, improving accuracy and reducing the cost of 13C-MFA experiments
Versatile coordination of iron dichloride with N-(fluoro)-aryl-substituted iminopyridine ligands: Synthesis, structures, magnetic properties and polymerization of isoprene
No full textInternational audienceAs iron is one of the most abundant resources on earth and a low toxicity transition metal, it is only natural that chemical reactions and catalysis revolving around this resource have attracted attention. Over the past three decades, considerable progress has been made in the field of coordination insertion polymerization using iron catalysts.[1] Of particular interest is a type of iron-based complexes supported by iminopyridine ligands, which enables the stereoselective cis or trans polymerization of conjugated dienes depending on the substituent on the N-imino group.[2] In this study, we focus on this particular family of complexes, more precisely on iminopyridine N-aryl fluorinated iron complexes {with iminopyridine 2-[(Ar)N=C(R)]C5H4N, where Ar = 3,5-(CF3)2C6H3 and R = H (L1) or CH3 (L3); Ar = C6H5 and R = H (L2) or CH3 (L4)}. Four complexes were isolated as single crystals, and their molecular structures were elucidated by X-ray diffraction, revealing distinct coordination depending on the nature of the ligand (Fig. 1). Mössbauer spectroscopy and magnetic susceptibility measurements confirmed the presence of Fe(II) in a high-spin state across all complexes, which was further supported by DFT calculations
Key Recovery from Side-Channel Power Analysis Attacks on Non-SIMD HQC Decryption
No full textInternational audienceHQC is a code-based cryptosystem that has recently been announced for standardization after the fourth round of the NIST post-quantum cryptography standardization process. During this process, the NIST specifically required submitters to provide two kinds of implementation: a reference one, meant to serve lisibility and compliance with the specifications; and an optimized one, aimed at showing the performance of the scheme alongside other desirable properties such as resilience against implementation misuse or side-channel analysis. While most side-channel attacks regarding PQC candidates running in this process were mounted over reference implementations, very few consider the optimized, allegedly side-channel resistant (at least, constant-time), implementations. Unfortunately, HQC optimized version only targets x86-64 with Single Instruction Multiple Data (SIMD) support, which reduces the code portability, especially for non-generalist computers. In this work, we present two power side-channel attacks on the optimized HQC implementation with just the SIMD support deactivated. We show that the power leaks enough information to recover the private key, assuming the adversary can ask the target to replay a legitimate decryption with the same inputs. Under this assumption, we first present a key-recovery attack targeting standard Instruction Set Architectures (ARM T32, RISC-V, x86-64) and compiler optimization levels. It is based on the well known Hamming Distance model of power consumption leakage, and exposes the key from a single oracle call. During execution on a real target, we show that a different leakage, stemming from to the micro-architecture, simplifies the recovery of the private key. This more direct second attack, succeeds with a 99% chance from 83 executions of the same legitimate decryption. While the weakness leveraged in this work seems quite devastating, we discuss simple yet effective and efficient countermeasures to prevent such a key-recovery
False Coverage Proportion Control for Conformal Prediction
No full textInternational audienceSplit Conformal Prediction (SCP) provides a computationally efficient way to construct confidence intervals in prediction problems. Notably, most of the theory built around SCP is focused on the single test point setting. In real-life, inference sets consist of multiple points, which raises the question of coverage guarantees for many points simultaneously. While on average, the False Coverage Proportion (FCP) remains controlled, it can fluctuate strongly around its mean, the False Coverage Rate (FCR). We observe that when a dataset is split multiple times, classical SCP may not control the FCP in a majority of the splits. We propose CoJER, a novel method that achieves sharp FCP control in probability for conformal prediction, based on a recent characterization of the distribution of conformal p-values in a transductive setting. This procedure incorporates an aggregation scheme which provides robustness with respect to modeling choices. We show through extensive real data experiments that CoJER provides FCP control while standard SCP does not. Furthermore, CoJER yields shorter intervals than the state-ofthe-art method for FCP control and only slightly larger intervals than standard SCP
50 ans de recherche en Ordonnancement - théorie et applications
No full textInternational audienceThis paper presents an overview of scheduling research done over the last half century. The main focus is on what is typically referred to as machine scheduling. The first section describes the general framework for machine scheduling models and introduces the notation. The second section discusses the basic deterministic machine scheduling models, including single machine, parallel machines, flow shops, job shops, and open shops. The third section describes more elaborate models, including multi-objective and multi-agent scheduling models, scheduling with controllable processing times, scheduling with rejection, just-in-time scheduling, scheduling with due date assignments, time-dependent scheduling, and scheduling with batching and setups. The two subsequent sections consider scheduling under uncertainty; section four goes into online and robust scheduling and section five covers stochastic scheduling models. The next section describes a variety of important scheduling applications, including applications in manufacturing, in services, and in information processing. The last section presents the main conclusions and discusses future research directions.Cet article présente un aperçu des recherches en ordonnancement menées au cours du dernier demi-siècle. L'accent est mis principalement sur ce que l'on appelle communément l'ordonnancement des machines. La première section décrit le cadre général des modèles d'ordonnancement des machines et introduit la notation. La deuxième section aborde les modèles d'ordonnancement déterministes de base, notamment les problèmes mono-machines, les problèmes à machines parallèles, et les problèmes d'atelier. La troisième section décrit des modèles plus élaborés, notamment les modèles d'ordonnancement multi-objectifs et multi-agents, l'ordonnancement avec temps de traitement contrôlables, l'ordonnancement avec rejet, l'ordonnancement juste-à-temps, l'ordonnancement avec attribution de dates d'échéance, l'ordonnancement dépendant du temps et l'ordonnancement avec traitement par lots et configurations. Les deux sections suivantes abordent l'ordonnancement en situation d'incertitude ; la quatrième section aborde l'ordonnancement en ligne et robuste, et la cinquième les modèles d'ordonnancement stochastique. La section suivante décrit diverses applications importantes de l'ordonnancement, notamment dans les secteurs de la fabrication, des services et du traitement de l'information. La dernière section présente les principales conclusions et aborde les orientations futures de la recherche