HAL-Université de Bretagne Occidentale
Not a member yet
75442 research outputs found
Sort by
Is AI Hitting the Storage Wall? and why We Should Care
No full textInternational audienceOne of the driving factors of good AI models is data volume (and quality). From the inference perspective, we are more and more using very complex and large models (or several simpler still large specialized models, e.g. for multi-agent and multi-model systems) that are supposed to fit into memory to run efficiently. Model size and use is growing faster than memory size available, forcing the system to rely on milli-second-latency storage systems. Are we about to hit the storage-wall soon
A low-complexity equalizer design for OTFS modulation in doubly-dispersive channels
No full textInternational audienceOrthogonal time frequency space (OTFS) modulation has emerged as a promising technique for reliable communication over rapidly time-varying channels, thanks to its robustness against Doppler effects and delay spread. However, the design of efficient equalizers for OTFS systems remains a key challenge due to the high computational complexity involved. In this paper, we propose a low-complexity linear equalizer for OTFS modulation that exploits the inherent structure of the effective channel matrix. Specifically, the proposed method leverages the block-circulant property and the sparse nature of the OTFS channel matrix and applies a Choleskybased decomposition to significantly reduce computational cost. Simulation results confirm that the proposed equalizer achieves notable improvements in both bit error rate performance and computational efficiency, outperforming existing state-of-the-art techniques
Diffusive gradient in thin film for ultra-trace methylmercury measurements in the coastal and open sea
No full textInternational audienceMonomethylmercury (MMHg) is a potent neurotoxin causing neurodevelopmental delays and cardiovascular and immunological issues. Human exposure primarily occurs through seafood consumption due to MMHg bioaccumulation and biomagnification from seawater into marine organisms. Determining MMHg in seawater at ultratrace concentrations poses logistical and analytical challenges. Diffusive Gradient in Thin-film (DGT) samplers represent a promising solution, which captures time-averaged concentrations by preconcentrating in situ MMHg over a defined exposure time. DGT manufactured with 3-mercaptopropyl-functionalized silica (3MFS) in agarose and polyacrylamide gels were tested and compared for the determination of MMHg present in open ocean and coastal waters. Different elution methods using acidic thiourea were tested to reach precise, accurate and quantitative elution of MMHg from the binding gel. We found that polyacrylamide-3MFS binding gels display a higher elution efficiency (94 ± 3 %), precision and better handling compared to agarose-3MFS gels (41 ± 6 %). A unique mooring line installed in the South Western Tropical Pacific Ocean, provided monthly DGT-MMHg concentrations over a year showing potential seasonal differences in MMHg concentrations ranging between 18 and 106 fM. DGT were also deployed in shallow Peruvian coastal waters, exhibiting higher MMHg concentrations (170 ± 97, n = 26) with typical benthopelagic gradients. DGT-MMHg concentrations were in good agreement with discrete water samples analyzed by reference methods using isotope dilution. DGTs offer complementary advantages over oceanographic cruises, notably in situ preconcentration, low blanks, minimal logistical requirements and cost-effectiveness. DGTs represent a valuable tool for studying the marine MMHg cycle for evaluating the implementation of the Minamata Convention
Replacing strain gages by line camera DIC in Hopkinson bar experiments
No full textInternational audienceStrain gages are widely used to acquire the signals in dynamic experiments with Hopkinson bars. Here, we explore the potential of displacement history measurements with line camera based digital image correlation (DIC) to substitute the role of strain gages and directly obtain particle velocity. After outlining the fundamental equations for deriving stress-strain curves, the technique is applied and validated through split-Hopkinson bar compression and tension tests, as well as direct impact experiments. In direct impact tests, the line camera enables simultaneous measurement of input and output forces, facilitating the verification of quasi-static equilibrium. Moreover, in cases where quasi-static equilibrium is clearly satisfied, a single line camera measurement on the striker bar is sufficient to determine the entire stress-strain curve. Compared to laser interferometry and photon Doppler velocimetry, the line camera DIC system demonstrates superior capability in measuring large displacements of Hopkinson bars. It also offers a reliable non-contact measurement alternative to strain gages, which are prone to delamination under high-impact conditions
Les inégalités territoriales d'éducation. Approches sociologiques et géographiques
No full textInternational audienc
Internal tide loss of coherence in a realistic simulation of the North Atlantic
No full textInternational audienceAbstract. The loss of coherence of the semidiurnal internal tide is investigated using a high-resolution realistic numerical simulation over the North Atlantic. The analysis focuses on processes resulting from the interaction between the internal tide and the mesoscale background flow at time scales typically shorter than one month. To this end, a theoretical framework based on vertical mode decomposition and the splitting of the internal tide signal into coherent and incoherent components is developed and applied to the outputs of the numerical simulation. This framework enables the transfer terms between the coherent and incoherent parts, and between the different vertical modes – and therefore horizontal scales – of the internal tides to be evaluated. By focusing on three subdomains with contrasting dynamics, we demonstrate that coherent-to-incoherent energy transfers significantly impact the internal tide energy budget. These transfers are dominated by advection by slowly varying flows and mainly occur without changing the vertical mode of the internal tide involved. This is attributed to the dominance of the barotropic and first baroclinic modes in the mesoscale flow combined with the structure of the mesoscale flow/internal tide interaction terms. Typical energy transfer rates are of the order of a few tens of days in the Gulf Stream region and a few hundred days in the Azores for the mode 1 internal tide
Enhancing Spatial Resolution of Sentinel-2 Imagery through Deep Learning and Generative Adversarial Networks: GS-SRGAN
No full textInternational audienceAbstract. Sentinel-2 satellites provide multi-spectral images with 13 bands at resolutions of 10, 20, and 60 m/pixel, widely used for various applications due to their cost-free access and high revisit frequency. Their open data policy has made them a key resource in remote sensing. Nonetheless, the growing need for high-resolution images has highlighted the significance of super-resolution technology (SR), which improves spatial detail through enhanced sensor precision and density. Deep learning techniques are an effective solution for enhancing Sentinel-2 images through super-resolution, improving low-resolution images by retrieving fine-grained high- frequency details. This results in high-resolution outputs from freely available data. In this research, we propose an enhancement of single-image resolution model derived from a Generative Adversarial Network, commonly abbreviated as GAN. We implemented and trained a model, named GS-SRGAN (Google Sentinel - SRGAN), built on the foundation of the Super-Resolution GAN model (SRGAN), using pairs of Google Earth and Sentinel-2 images for generating super-resolved outputs of the RGB bands from the multispectral Sentinel-2 data using a 4x scaling factor. The results from our GS-SRGAN model surpass those of current best in class models when evaluated using standard metrics such as SSIM (Structural Similarity Index) and PSNR (Peak Signal-to-Noise Ratio), enabling the super-resolved Sentinel-2 imagery for use in studies that demand very high spatial resolution
Recycling flax fibre composites by using a chemical solvolysis process in near- and supercritical conditions: linking fibre integrity to micro-mechanical performance
No full textAccords "Elsevier-Couperin"International audienceThe growing demand for bio-based composites has intensified the need for efficient recycling strategies that preserve fibre integrity and reduce environmental impact. In this study, unidirectional flax-reinforced composites with epoxy and poly-amide 11 (PA 11) matrices are manufactured and subsequently recycled using a solvolysis-recycling based process under near- and supercritical conditions in a semi-continuous flow reactor. A water/ethanol mixture (50/50 in mol) is employed as solvent at 250°C and 300°C under 250 bar. The recovered flax fibres are characterised using scanning electron microscopy (SEM), nanoindentation (NI) and atomic force microscopy (AFM). Mechanical tests on the original composites demonstrate tensile properties consistent with the best values reported in the literature, supported by low porosity levels (∼2%). After solvolysis, flax fibres recovered from PA 11 composites exhibit clean surfaces, preserved cell wall organisation, and nano-mechanical properties close to virgin fibres. In contrast, fibres from epoxy composites experience substantial degradation, especially at 300°C, with pronounced ultrastructural damage and a 65.8% reduction in indentation modulus. AFM confirms localised degradation patterns and lumen-driven damage pathways. These results highlight the superior recyclability of Flax – PA 11 systems and the intrinsic limitations of Flax – Epoxy composites for fibre-preserving solvolysis. Overall, the study provides composite manufacturers and recyclers with new insights into fibre-scale degradation mechanisms and supports the development of more circular, eco-efficient bio-composite architectures.</p
