HAL - Lille 3
Not a member yet
5967 research outputs found
Sort by
BAREM: A multimodal dataset of individuals interacting with an e-service platform
No full textInternational audienceThe use of e-service platforms has become essential for many applications (administrative documents, online shopping, reservations). Although these platforms have improved significantly the user experience, unexpected and stressful situations can occur. Navigation problems (latency, missing information, poor ergonomics) are not always reported to the designers. To address this problem, we propose a multimodal dataset (video, audio, and physiological data) to help implicitly quantify the impact of navigation problems on users when using an e-service platform. A scenario has been designed to generate various navigation problems which can lead to changes in user behaviour. A baseline is proposed to spot changes in user behaviour, opening the way towards automatically qualifying user experiences while using e-service platforms
Rapport d'évaluation. Étude de l'impact du programme "Prépa Compétences" sur l'accès à la formationdes demandeurs d'emploi dans le cadre du PIC
No full textA Study On the Effects of Pre-processing On Spatio-temporal Action Recognition Using Spiking Neural Networks Trained with STDP
No full textInternational audienceThere has been an increasing interest in spiking neural networks in recent years. SNNs are seen as hypothetical solutions for the bottlenecks of ANNs in pattern recognition, such as energy efficiency. But current methods such as ANN-to-SNN conversion and back-propagation do not take full advantage of these networks, and unsupervised methods have not yet reached a success comparable to advanced artificial neural networks. It is important to study the behavior of SNNs trained with unsupervised learning methods such as spiketiming dependent plasticity (STDP) on video classification tasks, including mechanisms to model motion information using spikes, as this information is critical for video understanding. This paper presents multiple methods of transposing temporal information into a static format, and then transforming the visual information into spikes using latency coding. These methods are paired with two types of temporal fusion known as early and late fusion, and are used to help the spiking neural network in capturing the spatio-temporal features from videos. In this paper, we rely on the network architecture of a convolutional spiking neural network trained with STDP, and we test the performance of this network when challenged with action recognition tasks. Understanding how a spiking neural network responds to different methods of movement extraction and representation can help reduce the performance gap between SNNs and ANNs. In this paper we show the effect of the similarity in the shape and speed of certain actions on action recognition with spiking neural networks, we also highlight the effectiveness of some methods compared to others
Déchiffrer la Société Géologique du Nord en escaladant les rayons de sa bibliothèque : histoire et analyse du fonds documentaire
No full textInternational audienceThe creation of the Société Géologique du Nord’s library is concomitant with the foundation of the society itself in 1870 and reflects its history, which is closely linked to that of the University of Lille. Library rules and organization, buildings and equipment, classification and cataloguing of documents: the archives provide us with a wealth of information on the day-to-day life of a learned society library and allow us to investigate the way in which this rich collection has been built up over the last 150 years, within an extraordinarily vibrant international network. Corresponding and associate members, relative learned societies have made it possible to build up an exceptional collection of documents through the exchange of publications: journals and periodicals focusing on geology, paleontology and the natural sciences from all over the world, books, -almost a quarter of which are not held in any other French university library-, brochures and manuscripts, the majority of which still have to be described and digitized. A full statistic survey analyses this collection both in its geographical and chronological aspects. Today, this collection, deposited and then donated to the University Library of Lille, is a jewel of the university heritage and still conceals many discoveries to come.La création de la bibliothèque de la SGN est concomitante de la fondation de la société en 1870 et reflète son histoire, étroitement mêlée à celle de l’Université de Lille. Règlement et fonctionnement pratique, locaux et aménagements, classement et catalogage des documents : les archives nous fournissent une mine d’informations sur le quotidien d’une bibliothèque de société savante et nous permettent d’investiguer la manière dont cette riche collection s’est constituée au fil des 150 dernières années, au sein d’un réseau international d’une extraordinaire vivacité. Membres correspondants et associés, sociétés savantes amies ont permis de construire par échanges de publications un ensemble documentaire exceptionnel : revues et périodiques de géologie, paléontologie et sciences naturelles provenant de tous les pays, ouvrages - dont près d’un quart ne sont conservés dans aucune autre bibliothèque universitaire française -, brochures et manuscrits, enfin, dont la plus grande partie reste à décrire et numériser. Une étude statistique complète permet d’analyser cette collection de manière géographique et chronologique. Aujourd’hui, ce fonds, déposé puis donné à la Bibliothèque Universitaire de Lille, est un fleuron du patrimoine universitaire et recèle encore bien des découvertes à venir
Conférence IIIF 2021 : session régionale IIIF France
No full textInternational audienceLe 24 juin 2021, l'Equipex Biblissima et la Bibliothèque nationale de France (BnF) ont co-organisé la session régionale IIIF France de la conférence internationale IIIF (International Image Interoperability Framework), qui s'est tenue du 22 au 24 juin 2021 (en ligne).La session a été co-animée par Jean-Philippe Moreux (BnF) et Régis Robineau (Biblissima). Elle a duré environ 2h et a donné l'opportunité à plusieurs projets et institutions de présenter leur activité et réalisations autour des standards et technologies IIIF
Poétique de la charade. Une expérience de la mouvance des morphèmes entre éclatement et jaillissement du signifiant
No full textInternational audienc
Familial adenomatous polyposis associated craniopharyngioma secondary to both germline and somatic mutations in the APC gene
No full textInternational audienc
Bayesian Optimization using Deep Gaussian Processes
No full textBayesian Optimization using Gaussian Processes is a popular approach to deal with the optimization of expensive black-box functions. However, because of the a priori on the stationarity of the covariance matrix of classic Gaussian Processes, this method may not be adapted for non-stationary functions involved in the optimization problem. To overcome this issue, a new Bayesian Optimization approach is proposed. It is based on Deep Gaussian Processes as surrogate models instead of classic Gaussian Processes. This modeling technique increases the power of representation to capture the non-stationarity by simply considering a functional composition of stationary Gaussian Processes, providing a multiple layer structure. This paper proposes a new algorithm for Global Optimization by coupling Deep Gaussian Processes and Bayesian Optimization. The specificities of this optimization method are discussed and highlighted with academic test cases. The performance of the proposed algorithm is assessed on analytical test cases and an aerospace design optimization problem and compared to the state-of-the-art stationary and non-stationary Bayesian Optimization approaches
