HAL - Université de Franche-Comté
Not a member yet
51444 research outputs found
Sort by
AGCD-Net: Attention Guided Context Debiasing Network for Emotion Recognition
No full textInternational audienceContext-aware emotion recognition (CAER) enhances affective computing in real-world scenarios, but traditional methods often suffer from context bias-spurious correlation between background context and emotion labels (e.g. associating “garden” with “happy”). In this paper, we propose AGCD-Net, an Attention Guided Context Debiasing model that introduces Hybrid ConvNeXt, a novel convolutional encoder that extends the ConvNeXt backbone by integrating Spatial Transformer Network and Squeeze-and-Excitation layers for enhanced feature recalibration. At the core of AGCD-Net is the Attention Guided - Causal Intervention Module (AG-CIM), which applies causal theory, perturbs context features, isolates spurious correlations, and performs an attention-driven correction guided by face features to mitigate context bias. Experimental results on the CAER-S dataset demonstrate the effectiveness of AGCD-Net, achieving state-of-the-art performance and highlighting the importance of causal debiasing for robust emotion recognition in complex settings
"Pour l’avantage des gens doctes" : la bibliothèque publique de Besançon et les bénédictins de Saint-Vanne (1694-1791)
No full textInternational audienc
A new dust map of the Milky Way: I. Principal features
No full textInternational audienceAims. We present REDLINE, a new technique for mapping three-dimensional interstellar extinction across the Galactic plane, with the aim of revealing the distribution of interstellar dust at distances where traditional mapping approaches face limitations, particularly beyond 4 kpc from the Sun. Methods. Our approach relies solely on near-infrared stellar observations to achieve a higher sensitivity through high-extinction regions. Building on our previous mapping techniques, REDLINE employs Markov chain Monte Carlo (MCMC) methods to constrain the line-of-sight dust distribution. A further methodological improvement is the incorporation of an enhanced Besançon Galaxy model (BGM), which provides more accurate stellar population predictions. We applied this refined framework to survey data covering the entire Galactic plane at |b| ≤ 1 • , resulting in a map with 10 ′ angular resolution but on a 5 ′ grid and with a 100 pc resolution in distance.Results. REDLINE successfully maps the structure of the Galactic disc, particularly in the inner Galaxy. While showing general agreement with existing maps of the large-scale dust distribution, our method provides superior sensitivity at greater distances compared to star-by-star analyses that excel at closer ranges. We identified several key spiral arm features and provide new constraints on dust distribution in previously poorly mapped regions. A significant finding from the REDLINE map indicates that dust in the Galactic disc is distributed in a highly non-uniform manner, revealing a fragmented structure with clumps of matter aligning along expected spiral patterns, rather than continuous spiral arms, and highlighting regions of reduced density that deviate from idealised logarithmic models. The seamless and extended view provided by REDLINE demonstrates its value for advancing understanding of the Milky Way's interstellar medium and spiral structure.</div
A multipoint perturbation formula for eigenvalue problems
No full textInternational audienceStandard perturbation theory of eigenvalue problems consists of obtaining approximations of eigenmodes in the neighborhood of a Hamiltonian where the corresponding eigenmode is known. Nevertheless, if the corresponding eigenmodes of several nearby Hamiltonians are known, standard perturbation theory cannot simultaneously use all this knowledge to provide a better approximation. We derive a formula enabling such an approximation result, and provide numerical examples for which this method is more competitive than standard perturbation theory
Searching for dark matter with a 1000 km baseline interferometer
No full textInternational audienceAxion-like particles (ALPs) arise from well-motivated extensions to the Standard Model and could account for dark matter. ALP dark matter would manifest as a nearly monochromatic field oscillating at an (as of yet) unknown frequency. The frequency depends on the ALP mass, which could plausibly range from eV/ to eV/. We report on a direct search for ALP dark matter through the ALP-nucleon interaction by interfering the signals of two atomic K-Rb-He comagnetometers, with one situated in Mainz, Germany, and the other in Kraków, Poland. We use the ALP dark matter's spatiotemporal coherence properties assuming the standard halo model of dark matter in the Milky Way to improve the sensitivity and exclude spurious candidates. The search extends over nine orders of magnitude in ALP mass. In this range, no significant evidence of an ALP signal is found. We thus place new upper limits on the ALP-neutron and ALP-proton couplings of GeV and GeV at a mass of eV/ and extending to a mass of eV/ where the upper limits reach below GeV and GeV, respectively. For both neutron and proton couplings, this work is an improvement of up to four orders of magnitude compared to previous laboratory constraints
Antibacterial Surfaces Prepared through Electropolymerization of N -Heterocyclic Carbene Complexes: A Pivotal Role of the Metal
No full textInternational audienceN-Heterocyclic carbene (NHC) complexes are known to have antibacterial properties in solutions. However, these complexes have never been immobilized on solid supports to prepare antibacterial surfaces. Here, we tackled this lack and succeeded in immobilizing these NHC complexes on gold surfaces by electropolymerization. For this, we synthesized a series of various NHC complexes of different low-valent transition metals (M = Ag(I), Au(I), Rh(I), Ru(II), Cu(I)) bearing a pyrrole function at the five-membered carbenic cycle. We measured the antibacterial properties of these complexes against two Gram-negative (Escherichia coli and Pseudomonas aeruginosa) and two Gram-positive bacteria (Staphylococcus aureus and Listeria innocua) by determining their minimum inhibitory concentration (MIC) values. All NHC complexes presented interesting antibacterial properties that are metal-dependent. The silver–NHC complex showed higher antibacterial activity against Gram-negative bacteria (MIC = 16 μg·mL–1) than against Gram-positive bacteria (MIC = 32 μg·mL–1) and was poorly efficient against L. innocua. All other metal–NHC complexes were more efficient against Gram-positive bacteria, with MIC values in the range 4–16 μg·mL–1. These NHC complexes were then electropolymerized on gold substrates using their pyrrole function. Efficient incorporation of these NHC species into polypyrrole (PPy) films was confirmed by X-ray photoelectron spectroscopy (XPS) measurements with metal contents ranging from 0.8% (Cu) to 12.3% (Ag). Scanning electron microscopy (SEM) and profilometry measurements ascertain that the homogeneity, structure, and thickness of the films depend on the metal. The antibacterial activities of the polypyrrole films were then determined by the halo inhibition method. A very good match between the antibacterial properties of the films and those of the monomers with Ag(I), Au(I), and Rh(I) complexes was found. For the other complexes, the metallic content was too low to obtain interesting antibacterial properties. The cytotoxicity of the films was finally evaluated on normal human dermal fibroblasts (NHDF). Our study reveals a strong impact of the doping anions of polypyrrole on cell viability
Identification of the conditions and minimum requirements necessary for the release of autologous fresh CAR T-cell products under hospital exemption: a position paper from the WP-bioproduction of the UNITC consortium
No full textInternational audienc
NUMERICAL SIMULATIONS FOR THE ARZ MODEL FOR VEHICULAR TRAFFIC WITH GENERAL POINT CONSTRAINTS ON THE DENSITY FLUX
No full textIn this paper we investigate numerically the model for vehicular traffic proposed in [B. Andreianov, C. Donadello, M.D. Rosini, A second order model for vehicular traffics with local point constraints on the flow, Mathematical Models and Methods in Applied Sciences 26 (4) (2016) 751-802].The model is obtained by coupling the ARZ system with point constraints on the density flux, and reproduces traffic behavior at bottlenecks. We implement an adapted version of the Glimm scheme featuring a non-classical Riemann solver at the constraint locations. We show its numerical convergence and its ability to correctly capture the evolution of both conservative and Riemann invariant variables in all situations, including those involving the vacuum.Our scheme allows to conduct a number of numerical experiments to prove that the model reproduces several typical phenomena such as capacity drop, faster is slower effect and phantom jams. Beyond the theoretical results obtained in the aforementioned paper, we also consider point constraints whose values are determined by a Lipschitz continuous function of time, or depend non-locally on the solution itself
Motility as a mediating variable in the influence of environmental concern on mobility habits
No full textInternational audienceThis study employs a second-order structural model (SEM) to conduct a mediation analysis, with the objective of elucidating the intricate relationship between environmental concerns and daily mobility habits. The mediation variable used is motility, defined as the personal capacity to be mobile, and combines individuals’ access, skills, and projects related to the transportation system. This study utilizes data from the second wave of the ‘National daily mobility panel’ (2019, ELIPSS) survey in France to demonstrate that higher environmental concern may lead to stronger public transportation habits and less frequent car use. However, the relationship between these variables is not straightforward but rather significantly contingent upon the unequal distribution of motility among the population. Among lower education or income groups, the lack of motility can impede the translation of environmental concern into mode shift. These findings contribute to the identification of more seamless and less standardized strategies for a transition towards socially and environmentally sustainable daily mobility habits