arXiv subjects

Gilles Roussel

Publications and source records attributed to Gilles Roussel.

6 recordsLinked to original sources

arXiv2026

A Latent Representation Learning Framework for Hyperspectral Image Emulation in Remote Sensing

Synthetic hyperspectral image (HSI) generation is essential for large-scale simulation, algorithm development, and mission design, yet traditional radiative transfer models remain computationally expensive and proposed emulation methods are often limited to spectrum-level outputs. In this work, we propose a latent representation-based framework for hyperspectral emulation that learns a probabilistic latent representation of hyperspectral data. The proposed approach supports both spectrum-level and spatial-spectral emulation and can be trained either in a direct one-step formulation or in a two-step strategy that couples variational autoencoder (VAE) pretraining with parameter-to-latent mapping. Experiments on PROSAIL-simulated vegetation data and Sentinel-3 OLCI imagery demonstrate that the method outperforms classical regression-based emulators in reconstruction accuracy, spectral fidelity, and robustness to real-world spatial variability. We further show that emulated HSIs preserve performance in downstream biophysical parameter retrieval, highlighting the practical relevance of emulated data for remote sensing applications.

cs.CV↗

arXiv2026

SVH-BD : Synthetic Vegetation Hyperspectral Benchmark Dataset for Emulation of Remote Sensing Images

This dataset provides a large collection of 10,915 synthetic hyperspectral image cubes paired with pixel-level vegetation trait maps, designed to support research in radiative transfer emulation, vegetation trait retrieval, and uncertainty quantification. Each hyperspectral cube contains 211 bands spanning 400--2500 nm at 10 nm resolution and a fixed spatial layout of 64 \times 64 pixels, offering continuous simulated surface reflectance spectra suitable for emulator development and machine-learning tasks requiring high spectral detail. Vegetation traits were derived by inverting Sentinel-2 Level-2A surface reflectance using a PROSAIL-based lookup-table approach, followed by forward PROSAIL simulations to generate hyperspectral reflectance under physically consistent canopy and illumination conditions. The dataset covers four ecologically diverse regions -- East Africa, Northern France, Eastern India, and Southern Spain -- and includes 5th and 95th percentile uncertainty maps as well as Sentinel-2 scene classification layers. This resource enables benchmarking of inversion methods, development of fast radiative transfer emulators, and studies of spectral--biophysical relationships under controlled yet realistic environmental variability.

cs.CV↗

arXiv2020

Gaussian Compression Stream: Principle and Preliminary Results

Random projections became popular tools to process big data. In particular, when applied to Nonnegative Matrix Factorization (NMF), it was shown that structured random projections were far more efficient than classical strategies based on Gaussian compression. However, they remain costly and might not fully benefit from recent fast random projection techniques. In this paper, we thus investigate an alternative to structured ran-om projections-named Gaussian compression stream-which (i) is based on Gaussian compressions only, (ii) can benefit from the above fast techniques, and (iii) is shown to be well-suited to NMF.

eess.SP↗

arXiv2018

Faster-than-fast NMF using random projections and Nesterov iterations

Random projections have been recently implemented in Nonnegative Matrix Factorization (NMF) to speed-up the NMF computations, with a negligible loss of performance. In this paper, we investigate the effects of such projections when the NMF technique uses the fast Nesterov gradient descent (NeNMF). We experimentally show the randomized subspace iteration to significantly speed-up NeNMF.

eess.SP↗

arXiv2016

D-LITe: Building Internet of Things Choreographies

In this work, we present a complete architecture for designing Internet of Things applications. While a main issue in this domain is the heterogeneity of Objects hardware, networks and protocols, we propose D-LITe, a solution to hide this wide range of low layer technologies. By abstracting the hardware, we focus on object's features and not on its real characteristics. D-LITe aims to give a universal access to object's internal processing and computational power. A small virtual machine embedded in each object gives a universal view of its functionalities. Each object's features are discovered and programmed through the network, without any physical access. D-LITe comes with the SALT language that describes the logical behaviour needed to include user's Objects into an IoT application. This communication is based on REST architecture. Gathering all these logical units into a global composition is our way to build a services Choreography, in which each Object has its own task to achieve. This paper presents also an analysis of the gain obtained when a Choreography is used instead of the most common services.

cs.SE↗

arXiv2015

A progressive mesh method for physical simulations using lattice Boltzmann method on single-node multi-gpu architectures

In this paper, a new progressive mesh algorithm is introduced in order to perform fast physical simulations by the use of a lattice Boltzmann method (LBM) on a single-node multi-GPU architecture. This algorithm is able to mesh automatically the simulation domain according to the propagation of fluids. This method can also be useful in order to perform various types of simulations on complex geometries. The use of this algorithm combined with the massive parallelism of GPUs allows to obtain very good performance in comparison with the static mesh method used in literature. Several simulations are shown in order to evaluate the algorithm.

cs.DC↗