Portail HAL des publications du LIRMM
Not a member yet
13279 research outputs found
Sort by
SRAM-based heavy ion beam flux and LET dosimetry
No full textInternational audienceThis paper explores the possibility of enhancing the capability of static random access memories (SRAMs) as heavy ion beam detectors starting from the multiple-cell upsets (MCUs) measured in some well characterized beams. In particular, the two main enablers brought by the MCU analysis are (1) the determination of the beam flux even when the LET of the beam is not known [whenever the LET is > 10 MeV/(mg/cm2)] and (2) the estimation of the LET of the heavy ion beam without reliance on any other instrument. The methods designed to determine these quantities are explained throughout the paper and are calibrated upon well characterized heavy ion beams. They are then put to test in less known heavy ion beams. Overall, the flux estimation, which exploits the saturation of the coverage, i.e., the ratio between MCU and beam fluence, instead of the unsaturated SEU cross section, can point out issues with beam calibration that can be corrected by the facility. The LET estimation, for which two different methods are proposed, when compared to Monte-Carlo simulations, showed a general agreement with an uncertainty of ∼3 MeV/(mg/cm2), which is acceptable for typical measurements in which the LET data-points are spread by a larger range
Moderate Exponential-time Quantum Dynamic Programming Across the Subsets for Scheduling Problems
No full textInternational audienceGrover Search is currently one of the main quantum algorithms leading to hybrid quantum-classical methods that reduce the worst-case time complexity for some combinatorial optimization problems. Specifically, the combination of Quantum Minimum Finding (obtained from Grover Search) with dynamic programming has proved particularly efficient in improving the complexity of NP-hard problems currently solved by classical dynamic programming. For these problems, the classical dynamic programming complexity in O, where O denotes that polynomial factors are ignored, can be reduced by a hybrid algorithm to O), with c < . In this paper, we provide a bounded-error hybrid algorithm that achieves such an improvement for a broad class of NP-hard single-machine scheduling problems for which we give a generic description. Moreover, we extend this algorithm to tackle the 3-machine flowshop problem. Our algorithm reduces the exponential-part complexity compared to the best-known classical algorithm, sometimes at the cost of an additional pseudo-polynomial factor
Secure protection of 3D content through reversible geometric deformation
No full textInternational audienceThe widespread use of 3D objects in many fields poses new challenges for their secure transmission, storage and visualization. While selective encryption methods allow format compliance and enable adjustable security by selectively encrypting 3D geometry, they produce visible noise and disrupt geometric coherence which reduces compression efficiency. In this paper, we propose a reversible deformation method for 3D object protection that breaks with state-of-the-art 3D security approaches by offering more natural and visually coherent protection. The deformation is applied directly to the geometry, controlled by a secret key, and can be adjusted to achieve the desired level of visual security. Experimental results presented on a large database demonstrate that our method offers protection comparable to selective encryption, while providing significantly greater resilience to reconstruction attacks such as Laplacian smoothing, particularly at low security levels where selective encryption is more vulnerable. This work introduces a new direction for format-compliant 3D object protection, designed for secure visualization in untrusted environments
Mapping biodiversity at very-high resolution in Europe
No full textSource Agritrop Cirad (https://agritrop.cirad.fr/616774/) * Autres projets (id;sigle;titre): 101060693;GUARDEN;(EU) safeGUARDing biodivErsity aNd critical ecosystem services across sectors and scales// 101060639;MAMBO;(EU) Modern Approaches to the Monitoring of BiOdiversity//International audienceThis paper describes a cascading multimodal pipeline for high-resolution biodiversity mapping across Europe, integrating species distribution modeling, biodiversity indicators, and habitat classification. The proposed pipeline first predicts species compositions using a deep-SDM, a multimodal model trained on remote sensing, climate time series, and species occurrence data at 50×50 m resolution. These predictions are then used to generate biodiversity indicator maps and classify habitats with Pl@ntBERT, a transformerbased LLM designed for species-to-habitat mapping. With this approach, continental-scale species distribution maps, biodiversity indicator maps, and habitat maps are produced, providing fine-grained ecological insights. Unlike traditional methods, this framework enables joint modeling of interspecies dependencies, bias-aware training with heterogeneous presence-absence data, and large-scale inference from multi-source remote sensing inputs
Archaeological heritage and biodiversity hotspots : mutualisation of technological innovation and investigation strategies to improve knowledge and protection in poorly explored deep-sea regions
No full textInternational audienceDeep-sea exploration is recent and its cultural and biodiversity heritage remains largely unknown. Human activities increasingly developing at great depths threaten the richest of these sites, and massive destruction occurs sometimes even before they are discovered and inventoried. Investigating and establishing effective protection measures for deep-sea biodiversity and cultural heritage sites is a critical challenge, requiring urgent and collaborative exploration efforts.Mutualistic approaches integrating deep-sea biodiversity and archeological investigations hold great potential to support this effort. Underwater archaeology and deep-sea ecology share similar requirements in terms of instrumentation for accessing sites, for observation and sampling at scales ranging from centimetres to tens-of-meters. Furthermore, the investigation of shipwrecks provides unique opportunities for advancing knowledge in terms of spatial and temporal dynamics of deep-sea communities and highlights the potential role of archaeological sites as habitats for protected species, of relevance to conservation strategies and to anthropogenic impact assessments at depth.Certain priority areas for underwater archaeology exploration, such as historical maritime routes, represent hotspots of this cultural and ecological heritage. This is the case of the Corsica Channel, between the east coast of the island and Italy, where the seabed reaches depths of several hundred meters within a short distance from the coast of Corsica, marked by a narrow shelf and steep margin, in addition to be host a large number of shipwrecks from more than a 2000-year time period. The development of « lightweight » deep-sea robotic technologies has opened access to these depths to explorers and researchers. The Département des Recherches Archéologiques Subaquatiques et Sous-marines (Drassm) of the French Ministry of Culture pioneered this effort through its programme of research and innovation on the deep-sea shipwrecks off-shore Corsica, fostering tight collaboration between archaeologists and roboticists for the development of dedicated instrumentation.This strategy and these tools are also highly valuable for deep-sea ecological research and biodiversity conservation programs. The development of collaborations between the Drassm and the Museum National d’Histoire Naturelle (Paris, France), since 2021, has notably provided new information on the distribution of protected taxa, such as scleractinian and gorgonian corals, in this region of the western Mediterranean sea under particular stress from climate disturbance. The data acquired provide insights to the growth rate and favourable habitat conditions of these foundation species and, more broadly, on the diverse fauna hosted by deep-sea shipwrecks, expanding deep-sea ecological knowledge, in alignment with the objectives of protection and conservation in the area (Parc Naturel Marin du Cap Corse et de l’Agriate, Natura 2000 Habitats Directive). This program triggered joined-efforts for the characterisation of settlements on different archaeological sites, their structuring species and associated fauna, using deep-sea imagery. The presentation will illustrate the value of this collaborative effort and its contribution to of the investigation of poorly explored sites, cultural and biological hotspot of unique richness, through different examples, ranging from the Roman-era to World War II shipwrecks
Pushing the frontiers of subexponential FPT time for Feedback Vertex Set
No full textThe paper deals with the Feedback Vertex Set problem parameterized by the solution size. Given a graph and a parameter , one has to decide if there is a set of at most vertices such that is acyclic. Assuming the Exponential Time Hypothesis, it is known that FVS cannot be solved in time in general graphs. To overcome this, many recent results considered FVS restricted to particular intersection graph classes and provided such algorithms. In this paper we provide generic conditions on a graph class for the existence of an algorithm solving FVS in subexponential FPT time, i.e. time , for some \varepsilon<1, where denotes the number of vertices of the instance and the parameter. On the one hand this result unifies algorithms that have been proposed over the years for several graph classes such as planar graphs, map graphs, unit-disk graphs, pseudo-disk graphs, and string graphs of bounded edge-degree. On the other hand it extends the tractability horizon of FVS to new classes that are not amenable to previously used techniques, in particular intersection graphs of ``thin'' objects like segment graphs or more generally -string graphs
End-to-end 3D instance segmentation of synthetic data and embryo microscopy images with a 3D Mask R-CNN
No full textInternational audienceIn recent years, the exploitation of three-dimensional (3D) data in deep learning has gained momentum despite its inherent challenges. The necessity of 3D approaches arises from the limitations of two-dimensional (2D) techniques when applied to 3D data due to the lack of global context. A critical task in medical and microscopy 3D image analysis is instance segmentation, which is inherently complex due to the need for accurately identifying and segmenting multiple object instances in an image. Here, we introduce a 3D adaptation of the Mask R-CNN, a powerful end-to-end network designed for instance segmentation. Our implementation adapts a widely used 2D TensorFlow Mask R-CNN by developing custom TensorFlow operations for 3D Non-Max Suppression and 3D Crop And Resize, facilitating efficient training and inference on 3D data. We validate our 3D Mask R-CNN on two experiences. The first experience uses a controlled environment of synthetic data with instances exhibiting a wide range of anisotropy and noise. Our model achieves good results while illustrating the limit of the 3D Mask R-CNN for the noisiest objects. Second, applying it to real-world data involving cell instance segmentation during the morphogenesis of the ascidian embryo Phallusia mammillata, we show that our 3D Mask R-CNN outperforms the state-of-theart method, achieving high recall and precision scores. The model preserves cell connectivity, which is crucial for applications in quantitative study. Our implementation is open source, ensuring reproducibility and facilitating further research in 3D deep learning
ClimBurst: A Novel Method to Detect Climatological Anomalies Over Time and Space
No full textInternational audienceDetecting abnormal climate events is crucial for understanding, predicting, and managing climate risks. However, most existing methods require prior knowledge about when and where to search for these events, limiting their effectiveness. In this study, we introduce ClimBurst, a new method to identify climate-related anomalies that does not require any prior information about their duration or spatial extent. We propose computing climate bursts to detect abnormal seasonal activity. The ClimBurst approach can detect anomalies at any time scale. The approach also compares anomalies at neighboring locations enabling the tracking of events across time and space. We apply our method on sea surface temperature data from the Mediterranean Sea between 1960 and 2021, where we detect particularly strong warm anomalies that can last from a few days to a few months over a few kilometers to hundreds, such as the 2015 marine heatwave. Our results reveal a noticeable increase in the frequency, magnitude and the spatial extent of these hot anomalies over time. Researchers and practitioners can use ClimBurst to detect and study climate anomalies, providing a basis for event attribution and long-term trend analysis
Bridging the gap between user stories and feature models by leveraging version control systems: A step towards software product line migration
No full textInternational audienceContext: Throughout the software lifecycle, a significant amount of knowledge is accumulated around the source code. In our work, we focus on agile software requirements, particularly user stories, and on issues and merge requests in version control systems, that have been opened for implementing user stories.Objective: The objective of this paper is to present a method that leverages this knowledge to guide an SPL migration. Methods: We consider merge requests in version control systems as the link between user stories (requirements) and the source code (implementation). The method combines Natural Language Processing (NLP) and clustering to identify features from user stories and hierarchically organize them. Relational Concept Analysis (RCA) is then used to compute logical rules from the hierarchy of features, using their links with the products and the source code. The logical rules are finally transformed into constraints in the produced feature model.Results: The method was implemented and evaluated on a dataset from an industrial partner. The results showed the efficiency of our method in synthesizing feature models for an SPL migration of the partner’s code base. Conclusion: The proposed method synthesizes feature models to guide an SPL migration based on agile software development practices and demonstrates its effectiveness on a real industrial dataset
Graph-Based Multitask Transfer Learning for Fault Detection and Diagnosis of Few-Shot Analog Circuits
No full textInternational audienceBuilding an interpretable fault detection and diagnostic model based on few-shot circuit samples and prior information about circuit structures is of significant importance. To fill these gaps, we propose a graph-based multitask transfer learning (TL) method for fault detection and diagnosis of circuits under few-shot conditions. First, in order to model the interconnections of nodes in a circuit, the sample data is organized into a graph structure, and a semi-supervised graphbased structural feature fusion method is proposed. The proposed method can accept graph-structured data and process the data using feature fusion methods. Second, to improve the model performance under few-shot conditions, two TL mechanisms are proposed for the topological structure characteristics of analog circuits as well as circuit signal characteristics. Finally, through a parameter-shared strategy, we propose a task transfer-based fault diagnosis approach. Experimental results on three different circuits show that the proposed method has the best diagnostic accuracy compared to typical detection and diagnosis schemes