A5004781743- Robotics and Sensor-Based Localization
- Nonlinear Optical Materials Studies
- Advanced Vision and Imaging
- Advanced Image and Video Retrieval Techniques
- Laser Material Processing Techniques
- Human Pose and Action Recognition
- Optical Coherence Tomography Applications
- Generative Adversarial Networks and Image Synthesis
- Optical Wireless Communication Technologies
- 3D Surveying and Cultural Heritage
- Metamaterials and Metasurfaces Applications
- GaN-based semiconductor devices and materials
- Photonic and Optical Devices
- Advanced Fiber Laser Technologies
- Advanced Antenna and Metasurface Technologies
- Advanced Image Processing Techniques
- Surface Modification and Superhydrophobicity
- Advanced Optical Imaging Technologies
- Advanced Optical Sensing Technologies
- ZnO doping and properties
- Ocular and Laser Science Research
- Plasmonic and Surface Plasmon Research
- Laser-induced spectroscopy and plasma
- Video Surveillance and Tracking Methods
- Laser-Matter Interactions and Applications
Université Côte d'Azur
2017-2024
Centre National de la Recherche Scientifique
2019-2024
Université de Bourgogne
2022-2024
Université Bourgogne Franche-Comté
2019-2022
Centre de Recherche sur l'Hétéro-Epitaxie et ses Applications
2021-2022
Laboratoire Interdisciplinaire Carnot de Bourgogne
2019-2022
Vision pour la Robotique
2022
Institut national de recherche en informatique et en automatique
2015-2021
Universidade Federal de Minas Gerais
2018-2021
Observatoire de la Côte d’Azur
2021
Abstract Deploying advanced imaging solutions to robotic and autonomous systems by mimicking human vision requires simultaneous acquisition of multiple fields views, named the peripheral fovea regions. Among 3D computer techniques, LiDAR is currently considered at industrial level for vision. Notwithstanding efforts on integration optimization, commercially available devices have slow frame rate low resolution, notably limited performance mechanical or solid-state deflection systems....
Semantic segmentation of images is an important issue for intelligent vehicles and mobile robotics because it offers basic information which can be used complex reasoning safe navigation. Different solutions have been proposed this problem along the last two decades, where recent deep neural networks approaches shown very promising results in context urban One main problems when comparing different semantic how to select appropriate metric evaluate their accuracy. On one hand, classic...
Hybrid components combining the optical power of a refractive and diffractive system can form compact doublet lenses that correct various aberrations. Unfortunately, diffraction efficiency these devices decreases as function deflection angle over element aperture. Here, we address this issue, compensating for chromatic dispersion correcting monochromatic aberrations with centimeter-scale hybrid-metalenses. We demonstrate correction at least 80% aberration 70% spherical aberration. finally...
Lidar, a technology at the heart of autonomous driving and robotic mobility, performs 3D imaging complex scene by measuring time flight returning light pulses. Many technological challenges, including enhancement observation field view (FoV), acceleration frame rate, improvement ambiguity range, reduction fabrication cost, component size, must be simultaneously addressed so that lidar reaches performance needed to strongly impact global market. We propose an innovative solution address...
<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> An approach to automatically detect impact craters on planetary surfaces is presented in this letter. It built up from a boosting algorithm proposed by Viola and Jones (2004) whose simplicity combined with an original learning strategy leads fast robust process consistent results. The validated image data sets Mars surface captured the Orbiter Camera onboard Global Surveyor probe. </para>
Polarization information of the light can provide rich cues for computer vision and scene understanding tasks, such as type material, pose, shape objects. With advent new cheap polarimetric sensors, this imaging modality is becoming accessible to a wider public solving problems, pose estimation, 3D reconstruction, underwater navigation, depth estimation. However, we observe several limitations regarding usage sensorial modality, well lack standards publicly available tools analyze...
Hot electrons generated within plasmonic structures possess a high kinetic energy that can be employed to drive and catalyze huge range of physicochemical processes at the metallic interface. Up now, these photogenerated hot carriers were mainly simple nanoparticles where carrier localization coincides spatially with position optical excitation. A current challenge for development future plasmonic-based electron devices requires ability delocalized production control on large-distance their...
In this paper, we propose a framework to build an improved metric representation of the environment with semantic information. We explore some recent advances deep neural networks object detection/semantic classification in visual-based perception scheme. The output system is map extended classes and their positioning. This combines sensors available commonly used mobile robotic platforms such as RGB-D camera, 2D LIDAR odometers. short, CNN-based detector 3D model-based segmentation...
Polarimetric imaging has been applied in a growing number of applications robotic vision (ex. underwater navigation, glare removal, de-hazing, object classification, and depth estimation). One can find on the market RGB Polarization cameras that capture both color polarimetric state light single snapshot. Due to sensor's characteristic dispersion, use lenses, it is crucial calibrate these types so as obtain correct polarization measurements. The calibration methods have developed far are...
Abstract The eyes of arthropods, such as those found in bees and dragonflies, are sophisticated 3D vision tools that composed an array directional microlenses. Despite the attempts achieving artificial panoramic by mimicking arthropod with curved microlens arrays, a wealth issues related to optical aberrations fabrication complexity have been reported. However, wide-angle imaging is becoming essential nowadays for autonomous robotic systems, yet most available solutions fail simultaneously...
We investigated the effect of temporal shaped femtosecond pulses on silicon laser micromachining. By using sinusoidal spectral phases, pulse trains composed sub-pulses with distinct separations were generated and applied to surface produce Laser Induced Periodic Surface Structures (LIPSS). The LIPSS obtained different sub-pulse separation analyzed by comparing intensity two-dimensional fast Fourier Transform (2D-FFT) AFM images ripples It was observed that amplitude is more emphasized for...
The development of advanced photonics devices requires materials with large optical nonlinearities, fast response times and high transparency, while at the same time allowing for micro/nano-processing needed integrated photonics. In this context, glasses have been receiving considerable attention given their relevant properties which can be specifically tailored by compositional control. Corning Gorilla® Glass (strengthened alkali aluminosilicate glass) is well-known its use as a protective...
ABSTRACT This article describes the design and fabrication of hierarchical nanomicrostructured polymer surfaces with high hydrophobicity. The nanoscale roughness is achieved by stamping a ZnO nanowire film into PDMS. Subsequently, microstructures different periodicities are created in stamped PDMS sample direct laser writing using femtosecond pulses. With this approach, we were able to produce surface morphologies, composed nano microscale structures that exhibit water contact angles larger...
This work describes the new facility for applied nuclear physics at University of Sao Paulo, mainly irradiation electronic devices. It is a setup composed quadrupole doublet beam focusing/defocusing plus multiple scattering through gold foils to produce low intensity, large-area, and high-uniformity heavy-ion beams from 1H 107Ag. Beam intensities can be easily adjusted 102 particles cm2/s hundreds nA an area as large 2.0 cm2 uniformity better than 90%. Its chamber has high-precision...
In this work, we performed the nonlinear characterization of Gorilla Glass waveguides produced by fs-laser microfabrication with different writing parameters. Their refractive index n2 was determined dispersive-scan technique. We found that values are lower than one for pristine material and they depend on Raman spectroscopy measurements showed important structural modifications related to non-bridging oxygens (NBOs) irradiated glass, which may lead reduction third-order polarizability.
We studied the near-band-edge emission (NBE) and yellow defect luminescence (YL) of GaN under below bandgap excitation with 40-fs laser pulses at a wavelength 775 nm. Even though in this case three-photon absorption processes are required for band-to-band transitions, fourth-order (INBE ∝ Iexc4) sub-linear (IYL Iexc0.5) dependencies were observed intensity NBE YL as function intensity, respectively. Modelling carrier dynamics few-level rate-equation model revealed that, high irradiances,...
Creating plausible virtual actors from images of real remains one the key challenges in computer vision and graphics. Marker-less human motion estimation shape modeling wild bring this challenge to fore. Although recent advances on view synthesis image-to-image translation, currently available formulations are limited transfer solely style do not take into account character's shape, which by nature intermingled produce forms. In paper, we propose a unifying formulation for transferring...
This paper proposes an environmental representation approach based on hybrid metric and topological maps as a key component for mobile robot navigation. Focus is made ego-centric pose graph structure by the use of Keyframes to capture local properties scene. With aim reducing data redundancy, suppress sensor noise whilst maintaining dense compact environment, neighbouring augmented spheres are fused in single representation. To this end, uncertainty error model propagation formulated outlier...