HAL Portal IOGS (nstitut d'Optique Graduate School)
Not a member yet
12589 research outputs found
Sort by
Advanced Traffic Engineering in WAN Using Graph Attention Networks
No full textInternational audienceEfficient and responsive traffic engineering is crucial for maintaining the robustness and reliability of Wide Area Networks (WANs). Traditional traffic engineering approaches often struggle to adapt to the dynamic and complex demands of today's network environments. To address these challenges, this paper enhances the Traffic Engineering algorithms by integrating an attention mechanism within the Edge-Path Embedding component. This significantly improves the model's adaptability and decision-making accuracy. Our comprehensive experimental evaluations demonstrate substantial improvements in terms of satisfied traffic demands and computational efficiency, highlighting the effectiveness of our approach
Excited-State Dynamics and Optical Properties of Silica Under Ultrafast Laser Irradiation
No full textInternational audienceExcited by intense infrared ultrafast light pulses, a wide bandgap material undergoes nonlinear ionization, generating a high density of free electrons in conduction states. As a result, the electronic band structure is critically modified and the bandgap shrinks. This induces rapid changes in opticalproperties, dramatically affecting the absorption spectrum during light coupling to the dielectric surface or during nonlinear propagation inside the bulk. This study analyzes the structural behavior and the modification of the optical properties of laser-excited silica glass at the molecular cluster level through first-principles simulations. Employing density functional theory and the GW approximations for band structure under nonequilibrium conditions, alongside the Bethe–Salpeter equation, the dynamics of the optical properties of fused silica are comprehensively explored. The behavior of excited fused silica in a wide photon energy range (from a few to 20 eV) is thus predicted. Laser-induced electron excitation triggers a redistribution of charges between oxygen and silicon atoms, accompanied by a significant increase in electronic pressure, local atomic structure rearrangement, and material expansion. Molecular dynamics simulations offer a temporal perspective on the excited state dynamics, unveiling the intricate interplay between electronic andatomic effects on bandgap evolution. The analysis sheds light on excitonic resonances, intraband and interband transitions in fused silica under ultrafast laser irradiation, providing valuable insights into its excited state behavior and optical properties
Insights into Photopolymerization at the Nanoscale Using Surface Plasmon Resonance Imaging
No full textNear-field photopolymerization (NFPP) driven by surface plasmon resonance has attracted increasing attention in nanofabrication. This interest comes from the nanometer-scale control of polymer thickness, due to the confinement of the evanescent wave within a highly restricted volume at the surface. In this study, a novel approach using a multi-spectral surface plasmon resonance instrument is presented that gives access to real-time images of polymer growth during NFPP with nanometer sensitivity. Using the plasmonic evanescent wave for both polymerization and real-time sensing, the influence of irradiance, concentration of dye, and initiator are investigated on the threshold energy and kinetics of NFPP. How oxygen inhibition in the near field strongly affects photopolymerization is highlighted, more than in the far field
Exploration de la dynamique de mouillage d'une nanogoutte d'eau sur des surfaces nano-texturées : perspectives atomistiques
No full textThe static and dynamic wetting behaviors of a water nanodroplet are investigated on flat and structured titanium (Ti) surfaces by using a fully atomistic simulation. Firstly, the results obtained for a water nanodroplet on a flat Ti surface agree with the static contact angle reported in the literature. Furthermore, to unveil the mechanisms of the water nanodroplet behavior on nano-textured surfaces, we analyze the impact of the roughness factors, the structure period, and sizes compared to the nanodroplet diameter. The introduction of the surface roughness produces an inhomogeneous spreading of the nanodroplet. The degree of roughness and the specific dimensions of the surface grooves determine these wetting processes. Surfaces with larger grooves and spacing tend to support more extensive spreading, while those with narrower grooves and higher roughness limit spreading. The obtained results also demonstrate damping in droplet vibrations accompanied by a gradual decrease in the dynamic contact angle of water nanodroplets on both flat and rough Ti surfaces, indicating a transition to a hydrophobic state. This effect is caused by the atomic bonding of water atoms to the surface ones and is shown to strongly depend on the surface roughness factors. The simulation results provide atomic level insights into the wetting mechanisms on surfaces featuring nanometric and sub-micrometric roughness, typically induced by femtosecond laser processing.Les comportements de mouillage statique et dynamique d'une nanogoutte d'eau sont étudiés sur des surfaces de titane (Ti) plates et structurées en utilisant une simulation entièrement atomistique. Tout d'abord, les résultats obtenus pour une nanogoutte d'eau sur une surface plate de Ti concordent avec l'angle de contact statique rapporté dans la littérature. De plus, afin de révéler les mécanismes du comportement de la nanogoutte d'eau sur des surfaces nano-texturées, nous analysons l'impact des facteurs de rugosité, de la période de la structure et des dimensions comparées au diamètre de la nanogoutte. L'introduction de la rugosité de la surface entraîne une répartition inhomogène de la nanogoutte. Le degré de rugosité et les dimensions spécifiques des rainures de la surface déterminent ces processus de mouillage. Les surfaces avec des rainures et des espacements plus larges tendent à favoriser une diffusion plus étendue, tandis que celles avec des rainures plus étroites et une rugosité plus élevée limitent la diffusion. Les résultats obtenus montrent également une atténuation des vibrations de la gouttelette accompagnée d'une diminution progressive de l'angle de contact dynamique des nanogouttes d'eau sur les surfaces de Ti plates et rugueuses, indiquant une transition vers un état hydrophobe. Cet effet est causé par la liaison atomique des atomes d'eau avec ceux de la surface et dépend fortement des facteurs de rugosité de la surface. Les résultats de la simulation fournissent des perspectives au niveau atomique sur les mécanismes de mouillage sur des surfaces présentant une rugosité nanométrique et submicrométrique, typiquement induite par le traitement au laser femtoseconde
Dirichlet process mixture model based on topologically augmented signal representation for clustering infant vocalizations
No full textBased on audio recordings made once a month during the first 12 months of a child's life, we propose a new method for clustering this set of vocalizations. We use a topologically augmented representation of the vocalizations, employing two persistence diagrams for each vocalization: one computed on the surface of its spectrogram and one on the Takens' embeddings of the vocalization. A synthetic persistent variable is derived for each diagram and added to the MFCCs (Mel-frequency cepstral coefficients). Using this representation, we fit a non-parametric Bayesian mixture model with a Dirichlet process prior to model the number of components. This procedure leads to a novel data-driven categorization of vocal productions. Our findings reveal the presence of 8 clusters of vocalizations, allowing us to compare their temporal distribution and acoustic profiles in the first 12 months of life
Generative shape deformation with optimal transport using learned transformations
No full textInternational audienceShape deformation is a fundamental problem in computer graphics and computer vision, with numerous appli- cations in fields such as animation, medical imaging, robotics to cite a few. We propose a method for shape deformation based on applying learned transformations with optimal transport (OT). Our method combines the power of the latter with the flexibility of learned transformations to provide an efficient and effective solution for 2D and 3D shape deformation. We formulate the problem as an OT task, where the goal is to learn the optimal way to move the mass distribution of a shape to another. We then use the learned geometric transformations, to achieve shape deformation. Our method can be applied to a wide range of shapes and applications. Interestingly, we show that it requires a small amount of data to learn the transformations. We demonstrate the performance of our method on our own crafted dataset of 2D and 3D shapes and evaluate its effectiveness using various metrics. The promising results obtained suggest that our method can be applied in a wide range of real-world applications
Enhancing a Many-Body Dipolar Rydberg Tweezer Array with Arbitrary Local Controls
No full textInternational audienceWe implement and characterize a protocol that enables arbitrary local controls in a dipolar atom array, where the degree of freedom is encoded in a pair of Rydberg states. Our approach relies on a combination of local addressing beams and global microwave fields. Using this method, we directly prepare two different types of three-atom entangled states, including a state and a state exhibiting finite chirality. We verify the nature of the underlying entanglement by performing quantum state tomography. Finally, leveraging our ability to measure multibasis, multibody observables, we explore the adiabatic preparation of low-energy states in a frustrated geometry consisting of a pair of triangular plaquettes. By using local addressing to tune the symmetry of the initial state, we demonstrate the ability to prepare correlated states distinguished only by correlations of their chirality (a fundamentally six-body observable). Our protocol is generic, allowing for rotations on arbitrary sub-groups of atoms within the array at arbitrary times during the experiment; this extends the scope of capabilities for quantum simulations of the dipolar XY model
Study of the lateral shift due to atmospheric refraction: alternative analytical methods, and new results
No full textInternational audienceAtmospheric refraction modifies the apparent position of objects in the sky, and also produces a progressive lateral shift of the light rays received from these objects; in the case of a spherically symmetric atmosphere, for the first time, this shift has been numerically studied in 2022, and different analytical estimators have been compared (by Labriji et al. ) for the total shift. This topic is important for the reconstruction of meteor trajectories, for the analysis of wavefront sensing in adaptative optics, etc. Always in the case of a spherically symmetric atmosphere, we show two other analytical methods to study this lateral shift, and to be able to estimate it analytically in the difficult case when the celestial object is seen near the astronomical horizon. One of these methods allows us to deduce an estimator, not only of the total shift, but also of the shift of any point of the ray. We compare properties of the total lateral shift and of the refraction angle, and also the chromatism of the total lateral shift to the chromatism of the air refractivity, for rays coming from an object seen either high enough above the astronomical horizon, or on it. In this latter case, our first method shows departures from proportionality between the chromatisms of the air refractivity, of the astronomical refraction angle, and, even more, of the total lateral shift
Distributed Optical Fiber-Based Radiation Detection Using an Ultra-Low-Loss Optical Fiber
No full textInternational audienceThe combination of an ultra-low-loss optical fiber sensitive to ionizing radiation and an optical time domain reflectometer (OTDR) is investigated to explore the feasibility of a single-ended distributed radiation detector. The peculiarity of the tested fiber resides in its regenerative high radiation-induced attenuation (RIA) response in the infrared spectrum (1310 nm), which returns to a low value once the irradiation has ended, combined to its sensitivity, highly increasing with the dose rate. In this work, only some sections of the fiber line were irradiated with 100 kV X-rays at room temperature, to prove the spatially resolved radiation detection capabilities of the system. The transient RIA response of the fiber was characterized at different pre-irradiation doses. A pre-irradiation treatment was shown to stabilize the optical fiber response, improving its RIA vs. dose rate linearity and repeatability. This improved response, in terms of radiation quantification, comes at the cost of a lower detection threshold. This work lays the bases for a distributed radiation detector, with some capabilities in dose rate evaluation
A Theoretically Grounded Extension of Universal Attacks from the Attacker's Viewpoint
No full textInternational audienceWe extend universal attacks by jointly learning a set of perturbations to choose from to maximize the chance of attacking deep neural network models. Specifically, we embrace the attacker's perspective and introduce a theoretical bound quantifying how much the universal perturbations are able to fool a given model on unseen examples. An extension to assert the transferability of universal attacks is also provided. To learn such perturbations, we devise an algorithmic solution with convergence guarantees under Lipschitz continuity assumptions. Moreover, we demonstrate how it can improve the performance of state-of-the-art gradient-based universal perturbation. As evidenced by our experiments, these novel universal perturbations result in more interpretable, diverse, and transferable attacks