HAL Université de Toulouse, et Toulouse INP
Not a member yet
159308 research outputs found
Sort by
High Temperature Operation and Spectral Stability of InGaN/GaN Ring Microlasers on Silicon
No full textInternational audienceIII-N ring microlasers on a silicon substrate with InGaN/GaN active layers emitting near 420 nm were investigated. The growth conditions and fabrication steps were optimized to realize stable lasing under optical pumping in cavities with a diameter of 6- 10 μ m. Chemically sensitive transmission electron microscopy images indicate that InGaN layers present in form of isolated islands. Between these InGaN islands, large areas of GaN are visible, forming barriers to lateral transport of free charge carriers in the active region and preventing their nonradiative surface recombination. For the first time, temperature stability of InGaN/GaN microring lasers characteristics are studied and lasing up to 100 degrees Celsius is demonstrated with the wavelength shift less than 1 nm. At room temperature, the threshold pump power is as low as 220 kW/cm 2 . The obtained results significantly expand the potential areas of application of III-N microlasers
Microphone array geometry-independent multi-talker distant ASR: NTT system for DASR task of the CHiME-8 challenge
No full textInternational audienceIn this paper, we introduce a multi-talker distant automatic speech recognition (DASR) system we designed for the DASR task 1 of the CHiME-8 challenge. Our system performs speaker counting, diarization, and ASR. It handles a variety of recording conditions, from dinner parties to professional meetings and from two speakers to eight. We perform diarization first, followed by speech enhancement, and then ASR as the challenge baseline. However, we introduced several key refinements. First, we derived a powerful speaker diarization relying on end-to-end speaker diarization with vector clustering (EEND-VC), multi-channel speaker counting using enhanced embeddings from EEND-VC, and target-speaker voice activity detection (TS-VAD). For speech enhancement, we introduced a novel microphone selection rule to better select the most relevant microphones among those distributed microphones and investigated improvements to beamforming. Finally, for ASR, we developed several models exploiting Whisper and WavLM speech foundation models. In this paper, we present the original results we submitted to the challenge and updated results we obtained afterward. Our strongest system achieves a 63% relative macro tcpWER improvement over the baseline and outperforms the challenge best results on the NOTSOFAR-1 meeting evaluation data among geometry-independent systems
Moving beyond metrics: Capturing the clinical context behind antibiotic prescriptions in French broiler production
No full textInternational audienceSignificant reductions in antimicrobial use (AMU) in food production animals have been observed over the last10 years across Europe. We sought to understand recent changes in AMU by characterising antibiotic prescribingpatterns in poultry production in the context of associated clinical information. We analysed trends in AMU forconventional broiler chicken production in France based on a dataset of 193,526 sales for 33,831 flocks on 2120farms for 2015–2023, including 21,218 antibiotic prescriptions. We found the percentage of flocks prescribedantibiotics dropped from 65 % in 2013 to 20 % in 2023, plateauing in 2020–2023 (oscillating between 13 % and23 %), and observed a reduction in the use of critical antibiotics. A multiple correspondence analysis and hierarchicalclustering on principal components of 1112 antibiotic prescriptions and associated clinical data for2021–2022 produced 1940 prescription events, grouped in five clusters of antibiotic prescribing patterns, eachcharacterised by a combination of clinical indicators related to age at treatment, lesions, syndromes, diagnoses,and isolated bacteria. Two main clusters were associated with bacterial diagnoses, suggesting that use of antibioticsin these clusters was necessary to manage disease. Two clusters were identified as potential targets forfurther interventions to improve antimicrobial stewardship, focusing on underlying factors driving AMU ratherthan outright reductions. Our findings raise questions about the sustainability of further reductions in AMU andtheir implications for animal health and welfare. This calls for a shift to a more sustainable approach to monitoringantimicrobial stewardship, using integrated indicators which consider AMU within its broader context
Role of the amine monomer in polymer/metal interaction: application to the DGEBA-DETA-TA6V system
No full textInternational audienc
Exploring the productivity per unit of livestock, land and labour of organic multi-species livestock farms in six European countries
No full textInternational audienceFarm diversification is increasingly suggested as a way to improve agriculture's productivity and sustainability. However, the role of livestock remains under-explored, particularly whether diversification of organic farms with multiple livestock species increases productivity. We assessed the livestock, land and labour productivity of 96 organic multi-species livestock farms in six European countries. We aggregated farm's production of commercialized crop and livestock products in a unit of agricultural production (kg of protein) and an economic unit (income in €). We then calculated agricultural productivity of each farm per unit of livestock (LSU), land (ha) and labour (AWU). We also calculated the livestock productivity (kg of protein per LSU) of each livestock enterprise (dairy cattle, beef cattle, meat sheep, dairy sheep, goats, poultry and pigs) on each farm. We found that most organic multi-species livestock farms are as productive as their specialised counterparts per unit of livestock, land and labour, with medians of 103.8 kg of protein/LSU, 91.2 kg of protein/ha and 2214.7 kg of protein/AWU, respectively. Farms that included a monogastric species had higher agricultural productivity than those that included ruminant species alone (i.e., cattle and sheep). The high variability among farms and livestock species requires nuancing the view of diversification as a silver-bullet strategy and exploring the factors that promote or hinder the success of livestock diversification to increase productivity. This study provides the first comparison of agricultural and economic productivity of organic multi-species commercial farms across six European countries and including seven livestock combinations
Motor complications and postural abnormalities interplay in Parkinson's disease
No full textInternational audienceBackground: Postural abnormalities (PA) and motor complications (MCs, including motor fluctuations - MFs- and levodopa-induced dyskinesia - LIDs) are hallmark of Parkinson's disease (PD) progression, yet their relationship remains poorly understood.Objective: To investigate the association between PA and MCs, motor symptoms, and non-motor symptoms (NMS) in patients with PD, and to assess whether PA influences the development of MCs over time.Methods: Data of the prospective NS-Park cohort (27 French PD Expert Centers) were analysed. PA was defined by a score ≥2 on item 3.13 of the MDS-UPDRS-III. Associations between PA and MCs, as well as with other motor symptoms and NMS, were assessed using logistic regression models. We used interval censoring survival models to assess the associations between PA at inclusion and the incidence of MCs. Analyses were adjusted for sex, age, disease duration, dopaminergic dose, and disease severity.Results: Among 13,037 included PD patients (58.7 % male, median age at diagnosis 61 years), 724 (5.6 %) presented with PA. Patients with PA had longer disease duration, higher disease severity, and higher dopaminergic treatment. PA exhibited a higher prevalence of troublesome MFs (OR: 5.96; 95 % CI: 4.25-8.32) and LIDs (OR: 2.81; 95 % CI: 1.79-4.30), while associations with milder MCs were inconsistent. However, PA was not significantly associated with the development of MCs during follow-up.Conclusions: PA are associated with more frequent severe MCs, and a higher burden of motor and NMS, making patient care particularly challenging
Foam formation in a gas–liquid stirred tank: Impact of impeller type and operating conditions
No full textInternational audienc
On L¹ and time-optimal state transitions in piecewise linear models of gene-regulatory networks
No full textInternational audienceIn this paper, we investigate optimal state transfers for a generic class of piecewise-linear models widely used to qualitatively describe gene-regulatory networks. Motivated by the main practical drawbacks of artificially regulating gene expression through chemical inducers, the optimality of the transitions is defined as the convex combination of the total time and the L¹ cost of the control. Solutions are studied through a Hybrid Pontryagin's Maximum Principle approach, which allows to characterize the optimal trajectories and control for the general formulation of the problem. Then, we focus on two practical examples of two-dimensional regulatory networks: the bistable switch, for which the objective is to induce optimal transitions between its two stable steady states, and the damped genetic oscillator, where the goal is to induce sustained oscillatory behaviors. The resulting optimal control strategies can be expressed in state feedback form, involving both bang arcs and inactive control periods, and are shown to slide over certain separatrices of the uncontrolled system that characterize the boundaries of the admissibility set
Two-level overlapping additive Schwarz preconditioner for training scientific machine learning applications
No full textInternational audienceWe introduce a novel two-level overlapping additive Schwarz preconditioner for accelerating the training of scientific machine learning applications. The design of the proposed preconditioner is motivated by the nonlinear two-level overlapping additive Schwarz preconditioner. The neural network parameters are decomposed into groups (subdomains) with overlapping regions. In addition, the network’s feed-forward structure is indirectly imposed through a novel subdomain-wise synchronization strategy and a coarse-level training step. Through a series of numerical experiments, which consider physicsinformed neural networks and operator learning approaches, we demonstrate that the proposed two-level preconditioner significantly speeds up the convergence of the standard (LBFGS) optimizer while also yielding more accurate machine learning models. Moreover, the devised preconditioner is designed to take advantage of model-parallel computations, which can further reduce the training tim