Portail HAL UHA (Université de Haute-Alsace)
Not a member yet
17715 research outputs found
Sort by
Fusion of Semantic Segmentation Models for Vehicle Perception Tasks
No full textInternational audienceIn self-navigation problems for autonomous vehicles, the variability of environmental conditions, complex scenes with vehicles and pedestrians, and the high-dimensional or real-time nature of tasks make segmentation challenging. Sensor fusion can representatively improve performances. Thus, this work highlights a late fusion concept used for semantic segmentation tasks in such perception systems. It is based on two approaches for merging information coming from two neural networks, one trained for camera data and one for LiDAR frames. The first approach involves fusing probabilities along with calculating partial conflicts and redistributing data. The second technique focuses on making individual decisions based on sources and fusing them later with weighted Shannon entropies. The two segmentation models are trained and evaluated on a particular KITTI semantic dataset. In the realm of multi-class segmentation tasks, the two fusion techniques are compared and evaluated with illustrative examples. Intersection over union metric and quality of decision are computed to assess the performance of each methodology
Focus last and agent first in basic varieties: what happens when the agent loses control?
No full textInternational audienc
Introduction : Traduire la littérature grand public et la vulgarisation
No full textInternational audienceIntroduction to the collected volume on Traduire la littérature grand public et la vulgarisation / Translating Popular Fiction and Science. Full-text made available from the publisher here: https://liseuse.harmattan.fr/979-10-309-0495-6Introduction au volume Traduire la littérature grand public et la vulgarisation / Translating Popular Fiction and Science. Article disponible en accès libre ici : https://liseuse.harmattan.fr/979-10-309-0495-
Phenothiazine-Carbazole-based Bis Oxime Esters (PCBOEs) for Visible Light Polymerization
No full textInternational audienceIn this study, a series of seven PCBOEs was innovatively designed and successfully synthesized for the first time by bridging a phenothiazine and a carbazole unit by mean of an acetylene spacer and bearing oxime ester groups as end groups on both sides. Structures of the different difunctional oxime esters were confirmed through NMR and HRMS. All PCBOEs exhibited excellent light absorption properties in the visible range. Upon excitation with a 405 nm LED for 60 s, a complete photolysis of the oxime esters could be achieved. Parameters such as the singlet state energy and the triplet state energy demonstrated the facile generation of the corresponding radicals upon light absorption by PCBOEs. Photopolymerization kinetics in TMPTA confirmed the homolytic cleavage of oxime esters and the generation of radicals. CO2 absorption peaks were detected when the PCBOEs/TMPTA systems were excited with a 405 nm LED. Based on the positive results, the photochemical mechanisms of polymerization with PCBOEs was proposed. PCBOE3 was identified as the best candidate of the series and this bifunctional oxime ester was successfully used for the 3D-printing of the “IS2M” logo via DLW. Experimental results from DSC also confirmed that PCBOEs not only serve as photoinitiators but also possess high reactivity as thermal initiator
Ultrafast sunlight-induced polymerization: Unveiling 2phenylnaphtho[2,3-d]thiazole-4,9-dione as a unique scaffold for high-speed and precision 3D printing
No full textInternational audienceA series of 15 dyes based on the 2‐phenylnaphtho[2,3‐ d ]thiazole‐4,9‐dione scaffold and 1 compound based on the 2,3‐diphenyl‐1,2,3,4‐tetrahydrobenzo[ g ]quinoxaline‐5,10‐dione scaffold are studied as photoinitiators. These compounds are used in two‐ and three‐component high‐performance photoinitiating systems for the free radical polymerization of trimethylolpropane triacrylate (TMPTA) and polyethylene glycol diacrylate (PEGDA) under sunlight. Remarkably, the conversion of TMPTA can reach ≈60% within 20 s, while PEGDA attains a 96% conversion within 90 s. To delve into the intricate chemical mechanisms governing the polymerization, an array of analytical techniques is employed. Specifically, UV–vis absorption and fluorescence spectroscopy, steady‐state photolysis, stability experiments, fluorescence quenching experiments, cyclic voltammetry, and electron spin resonance spin trapping (ESR‐ST) experiments, collectively contribute to a comprehensive understanding of the photochemical mechanisms. Photoinitiation capacities of these systems are determined using real‐time Fourier transformed infrared spectroscopy (RT‐FTIR). Of particular interest is the revelation that, owing to the superior initiation ability of these dyes, high‐resolution 3D patterns can be manufactured by direct laser write (DLW) technology and 3D printing. This underscores the efficient initiation of free radical polymerization processes by the newly developed dyes under both artificial and natural light sources, presenting an avenue for energy‐saving, and environmentally friendly polymerization conditions
Diagnostic cartographique des représentations d'un territoire nucléaire après la fermeture d'une centrale. Le cas de Fessenheim (France)
No full textInternational audienceThe closure of the nuclear power plan questions the notions of energy and territorial transitions. Could this event be the trigger for a reconversion of the territory? And, if so, how do residents imagine it? This contribution aims to reveal spontaneous representations of a nuclearized territory, to measure the place of nuclear power in discourse and its scope of impact. The protocol was constructed so that local residents were an integral part of the scientific activity by involving them in the production of knowledge on their territory. The objective is to test their ability to tackle a question – post-nuclear – in order to analyse their representations of the territory’s issues in a particular context, that of the closure of a single nuclear industry on which an entire commune has built its development. We are testing this hypothesis with local residents in order to access the propensity to go beyond nuclear identity and to project themselves, via their local practices, into another history (whether industrial or not). The approach is innovative in two ways: on the one hand, it is interested in the denuclearization process and, on the other hand, it uses the perception-based regional mapping (PBRM) non-directive method, in order to understand how citizens represent the challenges of the territory after the closure of the plant, and if nuclear power lastingly marks the identity of a territory. To do this, it involves conducting cartographic interviews to reveal the perceived issues, the spirit of the places experienced as experienced by the site inhabitants.La fermeture de la centrale nucléaire questionne profondément les notions de transitions énergétique et territoriale et leurs articulations. Cet évènement peut-il être le déclencheur d’une reconversion du territoire ? Et si oui, comment les habitants se le représentent-ils ?Cette contribution a pour objectif de faire apparaître les représentations spontanées d’un territoire nucléarisé, de mesurer la place du nucléaire dans les discours et son périmètre d’impact. Le protocole s’est construit de sorte que des riverains soient parties intégrantes de l’activité scientifique en les impliquant dans la production de connaissances sur leur territoire. L’objectif est de tester leur capacité à s’emparer d’une question – le postnucléaire – afin d’analyser leurs représentations des enjeux du territoire dans un contexte particulier, celui de la fermeture d’une mono-industrie nucléaire sur laquelle toute une commune a construit son développement. Nous testons cette hypothèse auprès d’habitants du territoire afin d’évaluer la propension à dépasser l’identité nucléaire et à se projeter, via leurs pratiques du territoire, dans une autre histoire (qu’elle soit industrielle ou non). L’approche est novatrice à double titre : d’une part, elle s’intéresse au processus de dénucléarisation et, d’autre part, elle utilise la méthode non- directive ZADA, Zonage A Dires d’Acteurs, afin de comprendre comment des citoyens, des « pratiquants d’un territoire nucléaire », se représentent les enjeux du territoire après la fermeture de la centrale, et si le nucléaire marque durablement l’identité d’un territoire. Pour ce faire, il s’agit de mener des entretiens cartographiques pour faire révéler les enjeux perçus, l’esprit des lieux vécus tels perçus par des habitants du site
A round-robin study on the tensile characterization of single fibres: A multifactorial analysis and recommendations for more reliable results
No full textInternational audienceIn this benchmark, the tensile properties of three types of organic fibres – flax, hemp and aramid − were determined using single-fibre tensile tests performed by nine research groups. Flax and hemp were chosen due to their prevalence among European fibre plants. Aramid was selected for its synthetic nature and comparable dimensional properties. Due to the morphological complexity and variability of plant fibres, the scatter in the apparent tangent modulus and strength is more pronounced for flax and hemp compared to aramid. The primary source of scatter in the tensile properties results from human factors and experimental procedures, particularly regarding the fibre selection, the measurement of the fibre cross-sectional area and of the tensile strain. The post-processing procedure also turns out to be a key factor. Finally, recommendations and guidelines for best practices are proposed to reduce the main sources of dispersion associated with the reproducibility of single fibre tensile tests
Efficiency of catalytic and non-catalytic ceramic foam for the filtration of particle emitted from a domestic wood stove
No full textInternational audienc
Deep Learning for Time Series Classification and Extrinsic Regression: A Current Survey
No full textInternational audienceTime Series Classification and Extrinsic Regression are important and challenging machine learning tasks. Deep learning has revolutionized natural language processing and computer vision and holds great promise in other fields such as time series analysis where the relevant features must often be abstracted from the raw data but are not known a priori. This article surveys the current state of the art in the fast-moving field of deep learning for time series classification and extrinsic regression. We review different network architectures and training methods used for these tasks and discuss the challenges and opportunities when applying deep learning to time series data. We also summarize two critical applications of time series classification and extrinsic regression, human activity recognition and satellite earth observation
Employing Singlet‐Singlet Energy Transfer for Boosting the Reactivity of Type I Photoinitiators in Radical Photopolymerization
No full textInternational audienceAbstract A remarkable and unexpected increase in the photopolymerization efficiency of an acrylic resin by a bisacylphosphine oxide photoinitiator was observed when an optical brightener was present in the medium. High values for the maximal rates of photopolymerization were obtained by RT‐FTIR at 365 nm under a very low irradiance of 1 mW/cm 2 . Fluorescence studies revealed that the quenching process occurs through singlet‐singlet energy transfer between the first singlet excited state of the optical brightener and the ground state photoinitiator. This mechanism acts as an additional pathway for the excitation of the photoinitiator, thereby increasing the total amount of initiating radicals. Using the Förster resonance energy transfer model, we calculated the relative efficiency of the photosensitization process with respect to the direct excitation efficiency of the photoinitiator. The results demonstrate that the photosensitization process can be predicted, paving the way for further improvements in photoinitiating systems