HAL Portal UTC Université de Technologie de Compiègne
Not a member yet
11652 research outputs found
Sort by
Rapport de Prospective Conseil Scientifique CNRS Ingénierie (Ex: INSIS)
No full textCe document est le rapport de prospective de CNRS Ingénierie, réalisé par le Conseil Scientifique de l'INSIS (CNRS Ingénierie) pour la mandature 2019-2023
Dépasser les polycrises
No full textCe 26e numéro de Rencontres transdisciplinaires aborde la persistance des crises mondiales, notamment celles liées à la COVID-19, la guerre, le climat, l’insécurité, les conditions sanitaires précaires et la diminution des ressources alimentaires. Des organisations telles que l’UNICEF, le FMI et la CNUCED alertent sur les multiples défis qui menacent la stabilité mondiale, avec le Secrétaire général de l’ONU, Antonio Guterres, soulignant la nécessité d’une action collective. Les auteurs posent des questions cruciales sur la façon de réagir à des crises sans frontières, mettant l’accent sur la nécessité d’une approche transdisciplinaire. Le texte propose dix contributions multilingues qui explorent diverses perspectives pour comprendre et dépasser ces crises, couvrant des thèmes tels que la relation humaine avec la nature, le changement climatique, l’inégalité des sexes et la promotion d’une culture de paix. Le numéro inclut également de nouvelles rubriques, notamment sur les activités des membres du CIRET et rend hommage à certains membres décédés dont Hubert Reeves
High Integrity Localization of Intelligent Vehicles with Student’s t Filtering and Fault Exclusion
No full textInternational audienceHigh-integrity localization is a key element for safety-critical applications like autonomous driving. The navigation filter plays a crucial role in merging sensor data to estimate an accurate pose and calculate a confidence interval based on task requirements. This paper presents an end-to-end Student's t information filter for accurate data fusion and non-pessimistic confidence domain computation. The filter incorporates a Fault Detection and Exclusion stage based on the Kullback-Leibler Divergence. The degree of freedom of the t distribution shapes the heavy tail to make the estimation process more robust against non detected outliers. We show that the adjustment of the degree of freedom can be done in real time using measurement residuals which give an indirect vision of the environment complexity.The accuracy and integrity of the proposed approach are evaluated with real data acquired with an experimental vehicle using GPS and Galileo pseudoranges merged with camera measurements after a map matching step with a High-Definition map. A comparative study with other classical methods based on Kalman filtering is also reported
Why Ethics is important to consider in AI ?
No full textInternational audienceEthics, defined initially as moral and virtues principles, becomes more and more deontological and social rules, especially to evaluate development and use of AI systems. In this paper, we present investigations to show the importance of ethics in AI technologies emphasizing their important impact in society and research
Stochastic Linear Quadratic Optimal Control of Speed and Position of Multiple Trains on a Single-Track Line
No full textInternational audienceIn the European Rail Traffic Management System (ERTMS), the Route Control Centre System (RCCS) supervises the distance between consecutive trains and generates movement authorities, i.e. the permission for a train to move to a specific location within the constraints of the infrastructure and with supervision of speed. In this work, a control model aimed at determining the speed and position of a train platoon within a sector of the rail network is presented. The central controller, i.e. the RCCS, receives information about the current position and speed of trains and it sends them decisions about optimal corrective actions for each train. Priorities of trains are handled to respect the planned timetable, taking into account the train dynamics, limitations in divergences of positions, speeds, and tractive effort, as well as minimum distances between consecutive trains. The control approach is based on a quite innovative linear quadratic regulator allowing the definition of stochastic constraints. The validation of the model is based on data collected from a RCCS for a Section of the high-speed Paris-London lin
Non-destructive Characterization of Skeletal Muscle Extracellular Matrix Morphology by Combining Optical Coherence Tomography (OCT) Imaging with Tissue Clearing
No full textInternational audienc
Les implicants premiers, un outil polyvalent pour l'explication de classification robuste
No full textNational audienceDans cet article, nous étudions comment les résultats d'un classifieur robuste peuvent être expliqués à l'aide d'implicants premiers, en nous concentrant sur l'explication de dominances par paires. Par robustes, nous sous-entendons des modèles prudents pouvant s'abstenir de classer ou de comparer deux classes lorsqu'ils manquent d'informations. Cela peut se faire en utilisant des ensembles (convexes) de probabilités. Par implicant premier, nous parlons d'un ensemble minimal d'attributs que nous devons figer afin d'obtenir une certaine conclusion (soit une dominance, soit une non-dominance entre deux classes). Après avoir présenté les concepts généraux, nous les appliquerons au cas bien connu du classifieur Bayesien naïf
Translating data science queries from natural language into graph analytics queries using NLDS-QL
No full textInternational audienceThis paper introduces NLDS-QL 1 , a translator of data science questions expressed in natural language (NL) into data science queries on graph databases. Our translator is based on a simplified NL described by a grammar that specifies sentences combining keywords to refer to operations on graphs with the vocabulary of the graph schema. This paper shows NLDS-QL in action within a scenario to explore and analyse a graph base with patient diagnoses generated with the open-source Synthea
SwimXYZ: A large-scale dataset of synthetic swimming motions and videos
No full textInternational audienceTechnologies play an increasingly important role in sports and become a real competitive advantage for the athletes who benefit from it. Among them, the use of motion capture is developing in various sports to optimize sporting gestures. Unfortunately, traditional motion capture systems are expensive and constraining. Recently developed computer vision-based approaches also struggle in certain sports, like swimming, due to the aquatic environment. One of the reasons for the gap in performance is the lack of labeled datasets with swimming videos. In an attempt to address this issue, we introduce SwimXYZ, a synthetic dataset of swimming motions and videos. SwimXYZ contains 3.4 million frames annotated with ground truth 2D and 3D joints, as well as 240 sequences of swimming motions in the SMPL parameters format. In addition to making this dataset publicly available, we present use cases for SwimXYZ in swimming stroke clustering and 2D pose estimation
