Archive Ouverte INSA Lyon
Not a member yet
75807 research outputs found
Sort by
Doppler-Supervised Channel Charting
No full textInternational audienceChannel charting is an integrated sensing and communication technique where passively collected channel state information (CSI) at the base-station is used for obtaining a low-dimensional representation of the distribution of all observed CSI samples, referred to as a channel chart. The usage of the channel chart for various applications depends highly on the charting accuracy, which is typically defined by its similarity to the users' geometric positions; an accurate channel chart can be used for a variety of applications like beam selection, beam tracking, radio maps, pilot assignment, signal strength prediction etc. Thus, improving the charting performance is of utmost importance. In this paper, we propose to exploit the delay-Doppler based multipath CSI which occurs naturally in modern techniques such as orthogonal time frequency space modulation as a form of supervision to produce channel charts that more closely match the geometric ground truth of user positions. We use a combination of classical triplet based loss function along with a novel Doppler based loss function to train a deep neural network which provides an efficient charting function. The charting performance for a variety of user trajectories demonstrates the effectiveness of our proposed approach
Unrolled Multiplicative Updates for Nonnegative Matrix Factorization applied to Hyperspectral Unmixing
No full textHyperSpectral Unmixing (HSU), the problem of separating mixed spectra of overlapping materials in a hyperspectral image, has motivated dedicated algorithmic developments in the last two decades. On the one hand, traditional model-based algorithms frequently guarantee interpretable results. On the other hand, deep-learning-based approaches are often faster at inference time and may obtain better empirical results. This work utilizes the strengths of both approaches by building on the deep unrolling paradigm. Our contribution is twofold. First, we propose two new algorithms based on deep unrolling of the well-known Multiplicative Updates. The first, coined Non-Adaptive Learned Multiplicative Updates (NALMU), adopts a simple element-wise multiplicative scheme. The second, called Recursive Adaptive Learned Multiplicative Updates (RALMU), has more flexible updates and better take into account the spatial correlations in the abundances. Second, we relate NALMU to the minimization of an explicit cost function under some assumptions. Such guarantees are unique in the HSU field. NALMU and RALMU are tested on astrophysics and remote sensing datasets. They outperform the other deep learning-based HSU algorithms and classical iterative schemes for the endmember estimates and obtain competitive results for the abundance estimates, even when trained in a self-supervised way. The code used in this paper will be made available upon publication
Sobolev estimates for the Keller-Segel system and applications to the JKO scheme
No full textWe prove L^{\infty}_{t} W^{1,p} Sobolev estimates in the Keller-Segel system by proving a functional inequality, inspired by the Brezis-Gallouët-Wainger inequality. These estimates are also valid at the discrete level in the Jordan-Kinderlehrer-Otto (JKO) scheme. By coupling this result with the diffusion properties of a functional according to Bakry-Emery theory, we deduce the L^{2}_{t} H^{2}_{x} convergence of the scheme, thereby extending the recent result of Santambrogio and Toshpulatov in the context of the Fokker-Planck equation to the Keller-Segel system
Training for Mixed-Precision Integer Weights, Activations and Embeddings in BERT
No full textInternational audiencePre-trained language models (LMs) deliver strong performance across a wide range of Natural Language Processing (NLP) tasks but remain costly to deploy on embedded devices due to their high memory and compute requirements. A widely used strategy for adapting LMs to resource-constrained devices is aggressive quantization. At low bit-widths, mixed-precision schemes, where different components of the model use different numerical precisions, offer an effective balance between compression and accuracy. In this work, we evaluate the impact of mixed-precision quantization for inference on the BERT language model. Unlike prior studies that often overlook activation quantization, our evaluation systematically explores mixed-precision configurations for both weights and activations. We also examine the effects of quantizing the embedding layer, which is commonly limited to token-weight quantization. Evaluated on the SQuAD and GLUE benchmarks, our approach achieves substantial reductions in memory and computational cost without sacrificing accuracy
Liposome‐Based Potential Vaccines Platforms that Are Noncytotoxic
No full textInternational audienceSynthetic vaccines represent a promising avenue in cancer immunotherapy by promoting targeted immune responses. Liposomal technologies have further advanced synthetic vaccinology by enabling the efficient delivery of tumor‐associated carbohydrate antigens. Despite this progress, the toxicity and reproducibility of such platforms remain underexplored. In this preliminary study, we synthesized a series of neoglycolipids bearing the Thomsen–Nouveau (Tn) antigen using bio‐orthogonal thiol–ene click chemistry. Here we present the results obtained using a set of neoglycolipids that were evaluated for their ability to self‐assemble into liposomal vesicles and for in vitro cytotoxicity. The resulting neoglycolipids exhibited no detectable cytotoxicity and formed stable liposomal structures when formulated with palmitic acid and 1‐palmitoyl‐2‐oleoyl‐ sn ‐glycero‐3‐phosphocholine via a freeze–thaw/extrusion process. This early‐stage work presents a proof of concept for a tunable, liposomal‐based synthetic vaccine platform
Coupler les données satellitaires aux données de terrain pour comprendre les dynamiques des socio-hydrosystèmes à différents niveaux d'organisation
No full textNote scientifique du PEPR OneWate
Integrated modelling of sediment and organic carbon fluxes in a large catchment: quantifying riverine contributions to the Mediterranean Sea
No full textInternational audienceWithin the Land Ocean Aquatic Continuum, the export of dissolved and particulate organic carbon (DOC, POC) is crucial for various processes in aquatic ecosystems and a primary source of greenhouse gas emissions in hydrosystems. Quantifying DOC and POC fluxes at high spatiotemporal resolution is complex because numerous sources and processes influence their dynamics, including climate, geological, and anthropogenic activities. Such quantifications at the catchment scale can be performed using hydrological models that consider OC processes. This study employed the Soil and Water Assessment Tool -Carbon (SWAT-C) to simulate OC dynamics in the large Mediterranean catchment of the Rhône River from 2002 to 2020. We represent hydrological and sediment fluxes at the catchment scale by accounting for sub-catchment diversity and dam operations. The average resulting exports reach 5.52 Mt y -1 , 78.8 kt y -1 , and 95.1 kt y -1 with 54.4%, 71.5%, and 44.3% exported during the highest discharges for sediment, POC, and DOC, respectively. The contributions of the southern sub-basins to POC and DOC fluxes are significant, while the northern sub-basins do not show significant patterns. Regarding OC quality, the proportions of labile/refractory OC at the Rhône outlet are 85/15 and 5/95 for POC and DOC, respectively. Global changes could alter flows, such as the labile/refractory OC proportions, which should be quantified using the tool developed in this article to anticipate their impacts on biogeochemical dynamics in the Mediterranean Sea. Finally, our study demonstrates the potential and limitations of SWAT-C for modelling OC dynamics in a large Mediterranean catchment under multiple anthropogenic pressures
k-scale: k-Anonymizing Millions of Trajectories
No full textInternational audienceTrajectory datasets collected by network operators and service providers offer detailed information about individual mobility and have wide application in business and research. However, managing such data raises privacy risks, as the unique movement patterns of individuals pose significant re-identification risks and make common countermeasures like pseudonymization ineffective. The privacy-preserving data publishing (PPDP) of trajectory datasets that maintains post-anonymization accuracy and truthfulness is an open problem -especially for large datasets with millions of records like those gathered by major actors in the telco ecosystem. We close this gap with k-scale, a framework that implements k-anonymity in massive mobile user trajectory datasets, removing uniqueness while safeguarding accuracy at the record level. Not only k-scale is the first model capable of scaling k-anonymization to a dataset of one million trajectories, but it does so while also outperforming state-of-the-art methods for trajectory data publishing in terms of preserved data quality, which we prove in real-world massive datasets and applications
Enhancing data anomaly prediction and real-time physical problem detection with Digital Twins and Cognitive Super Digital Twins
No full textInternational audienceThe increasing reliance on Internet of Things (IoT) systems has highlighted the need for effective strategies to detect data anomalies and address physical malfunctions in real time. This paper proposes a Cognitive Super Digital Twin (CSDT) to detect data anomalies and flag physical problems in IoT systems. The framework augments a standard DT with a synthetic data layer that balances rare events and improves model learning. We validate the method on an environmental-sensing and robot actuation case study, showing higher recall and F1 when training with augmented data, and a practical DT that detects incomplete robot motions. The approach improves resilience while keeping costs low by prioritizing digital experiments before physical changes
pELECTRE_Tri: probabilistic multiple-criteria decision-making method for sorting and classification
No full textA computational framework and a Python 3.11.7 implementation of ELECTRE Tri-B multiple-criteria decision-making and of Simos' method for determining the weights of criteria, along with documentation and examples