A5020881280- Advanced Fluorescence Microscopy Techniques
- Image and Signal Denoising Methods
- Advanced Image Processing Techniques
- Calibration and Measurement Techniques
- Advanced Image Fusion Techniques
- Adaptive optics and wavefront sensing
- Image Processing Techniques and Applications
- Sparse and Compressive Sensing Techniques
- Galaxies: Formation, Evolution, Phenomena
- Photoacoustic and Ultrasonic Imaging
- Optical Coherence Tomography Applications
- Business Strategy and Innovation
- Radio Astronomy Observations and Technology
- International Science and Diplomacy
- Astrophysics and Star Formation Studies
- Statistical Methods and Inference
- Stellar, planetary, and galactic studies
- Remote-Sensing Image Classification
- Parallel Computing and Optimization Techniques
- Medical Image Segmentation Techniques
- Infrared Target Detection Methodologies
- Spectroscopy and Chemometric Analyses
- Advanced Vision and Imaging
- Cosmology and Gravitation Theories
- Statistical and numerical algorithms
Centre National de la Recherche Scientifique
2012-2025
Université Paris-Saclay
2016-2025
Laboratoire des signaux et systèmes
2016-2025
CentraleSupélec
2016-2021
Université Paris-Sud
2010-2019
Institut d'Astrophysique Spatiale
2016-2019
Institut d'Astrophysique de Paris
2007-2015
Supélec
2009-2012
Sorbonne Université
2012
Institut Pasteur
2011
Structured illumination microscopy is a recent imaging technique that aims at going beyond the classical optical resolution by reconstructing high-resolution (HR) images from low-resolution (LR) acquired through modulation of transfer function microscope. The implementation has number drawbacks, such as requiring large to be and parameters manually set in an ad-hoc manner have, until now, hampered its wide dissemination. Here, we present new framework based on Bayesian inverse problem...
This paper tackles the problem of image deconvolution with joint estimation PSF parameters and hyperparameters. Within a Bayesian framework, solution is inferred via global posteriori law for unknown object. The estimate chosen as posterior mean, numerically calculated by means Monte-Carlo Markov chain algorithm. estimates are efficiently computed in Fourier domain effectiveness method shown on simulated examples. Results show precise hyperparameters well including restoration...
Context. Extracting the multiphase structure of neutral interstellar medium is key to understanding star formation in galaxies. The radiative condensation diffuse warm producing a thermally unstable lukewarm and dense cold closely related initial step leading atomic-to-molecular (HI-to-H 2 ) transition molecular clouds. Up now, mapping these phases out 21 cm emission hyper-spectral cubes has remained elusive mostly due velocity blending individual structures present on given line sight. As...
This paper is devoted to the problem of sampling Gaussian distributions in high dimension. Solutions exist for two specific structures inverse covariance: sparse and circulant. The proposed algorithm valid a more general case especially as it emerges linear problems well some hierarchical or latent models. It relies on perturbation-optimization principle: adequate stochastic perturbation criterion optimization perturbed criterion. proved that optimizer sample target distribution. main...
Terahertz Time Domain Spectroscopy (THz-TDS) systems have emerged as mature technologies with significant potential across various research fields and industries. However, the lack of standardized methods for signal noise estimation reduction hinders its full potential. This paper introduces a methodology to significantly reduce in THz-TDS time traces, providing reliable less biased signal. The method results an improved signal-to-noise ratio, enabling utilization dynamic range such setups....
We present a robust method, as well new metric, for the comparison of permittivity models in terahertz timedomain spectroscopy (THz-TDS).In this work, we show that signal is perturbed by deterministic part has to be either modeled or removed.Therefore, perform an extensive perturbation/noise analysis THz-TDS system, remove and model unwanted perturbation/noises implement them into our fitting process.This method now been implemented open-source software, Fit@TDS, available at :...
Context: Sh2-104 is a Galactic H ii region with bubble morphology, detected at optical and radio wavelengths. It considered the first observational confirmation of collect-and-collapse model triggered star-formation. Aims: We aim to analyze dust gas properties better constrain its effect on local future generations stars. In addition, we investigate relationship between emissivity index {\beta} temperature, T_dust. Methods: Using Herschel PACS SPIRE images 100, 160, 250, 350 500 {\mu}m...
We present an original method for reconstructing a 3-D object having two spatial dimensions and one spectral dimension from data provided by the infrared slit spectrograph on board Spitzer Space Telescope. During acquisition, light flux is deformed complex process comprising four main elements (the telescope aperture, slit, diffraction grating, optical distortion) before it reaches 2-D sensor. The originality of this work lies in physical modeling, integral form, formation continuous...
This paper deals with Gibbs samplers that include high dimensional conditional Gaussian distributions. It proposes an efficient algorithm avoids the sampling and relies on a random excursion along small set of directions. The is proved to converge, i.e., drawn samples are asymptotically distributed according target distribution. Our main motivation in inverse problems related general linear observation models their solution hierarchical Bayesian framework implemented through algorithms....
In this work, we introduce an original strategy to apply the Compressed Sensing (CS) framework a super-resolution Structured Illumination Microscopy (SIM) technique. We first define for direct domain CS, that exploits sparsity of fluorescence microscopy images in Fourier domain. then propose application method fast 4-images SIM technique, which allows reconstruct super-resolved using only 25% camera pixels each acquisition.
Hyperspectral imaging (HSI) has become an invaluable tool for many applications in astrophysics or Earth observation. Unfortunately, direct observation of hyperspectral images is impossible since the actual measurements are 2-D and suffer from strong spatial spectral degradations, especially infrared. We present this work original method high-resolution image reconstruction heterogeneous degraded by integral field spectroscopy (IFS) instrument. A fundamental part developing a forward model...
We investigate super-resolution methods for image reconstruction from data provided by a family of scanning instruments like the Herschel observatory. To do this, we constructed model instrument that faithfully reflects physical reality, accurately taking acquisition process into account to explain in reliable manner. The inversion, ie process, is based on linear approach resulting quadratic regularized criterion and numerical optimization tools. application concerns maps SPIRE evaluation...
We describe regularized methods for image reconstruction and focus on the question of hyperparameter instrument parameter estimation, i.e. unsupervised myopic problems. developed a Bayesian framework that is based posteriordensity all unknown quantities, given observations. This density explored by Markov chain Monte-Carlo sampling technique Gibbs loop including Metropolis-Hastings step. The numerical evaluation relies SPIRE Herschel observatory. Using simulated real observations, we show...
We propose a structured illumination microscopy method to combine super resolution and optical sectioning in three-dimensional (3D) samples that allows the use of two-dimensional (2D) data processing. Indeed, obtaining super-resolution images thick is difficult task if low spatial frequencies are present in-focus section sample, as these have be distinguished from out-of-focus background. A rigorous treatment would require 3D reconstruction whole sample using point spread function stack...
Astrophysical images issued from different instruments and/or spectral bands often require to be processed together, either for fitting or comparison purposes. However each image is affected by an instrumental response, also known as point-spread function (PSF), that depends on the characteristics of instrument well wavelength and observing strategy. Given knowledge PSF in band, a straightforward way processing homogenise them all target using convolution kernels, so they appear if had been...
Super-resolution in structured illumination microscopy (SIM) is obtained through de-aliasing of modulated raw images, which high frequencies are measured indirectly inside the optical transfer function. Usual approaches that use 9 or 15 images often too slow for dynamic studies. Moreover, as experimental conditions change with time, modulation parameters must be estimated within images. This paper tackles problem image reconstruction fast super resolution SIM, where number available reduced...
The use of IEEE 754-2008 half-precision floating-point numbers is an emerging trend in Graphical Processing Units' architecture. Being such a compact way representing data, its may speed up programs by reducing the memory bandwidth usage and allowing hardware designers to fit more computing units within same die space. In this paper, we highlight acceleration offered half over different implementations operation, 2D convolution. We show that even though it lead significant speed-up,...
In this paper we propose an approach of image restoration from multispectral data provided by imaging system. We specifically address two topics: (i) Development a multi-wavelength direct model for non-stationary instrument response that includes spatial convolution and spectral integration, (ii) Implementation using regularized least-square, based on quadratic criterion minimized gradient algorithm. test our simulated the Mid-InfraRed Instrument IMager (MIRIM) James Webb Space Telescope...
Energy efficiency is key in many embedded systems order to reach the best performance on a limited power budget. In addition, new applications based neural networks integrate various processing requirements, leading use of dedicated hardware functions optimize energy efficiency. Heterogeneous system-on-chips (SoC) bring together different computing capabilities, such as Nvidia Jetson AGX Orin. This type SoC includes CPU for general-purpose processing, GPU intensive data parallelism, and Deep...
Structured Illumination Microscopy (SIM) is an imaging technique for achieving both super-resolution (SR) and optical sectioning (OS) in wide-field microscopy. It consists illuminating the sample with periodic patterns at different orientations positions. The resulting images are then processed to reconstruct observed object SR and/or OS. In this work, we present BOSSA-SIM, a general-purpose SIM reconstruction method, applicable moving objects such as encountered vivo retinal imaging, that...