Scientific Publications of the University of Toulouse II Le Mirail
Not a member yet
92205 research outputs found
Sort by
Fifty years of research on resource-constrained project scheduling explored from different perspectives
No full textInternational audienceThe resource-constrained project scheduling problem is one of the most investigated problems in the project scheduling literature, and has a rich history. This article provides a perspective on this challenging scheduling problem, without having the ambition to provide a complete overview. Instead, the article does aim to summarize a number of reasons why this problem has been so intensely investigated from different perspectives.It will be shown that this scheduling problem has many faces, and therefore deserves a lot of research time from a computational and theoretical point of view as well as from a practical point of view. An overview of possible extensions to other problems and a detailed overview of the used (both heuristic and exact) solution methods will be given. In addition, the data used will be discussed and interesting avenues for further research will be mentioned throughout the different sections.</div
Les lieux que crée l’exil : les migrants russes à Tbilissi, 2022-2024
No full textInternational audienceDepuis l’invasion russe de l’Ukraine en 2022, de nombreux migrants russes se sont installés à Tbilissi, Géorgie, créant des espaces associatifs et des « tiers-lieux » pour reconstruire leurs réseaux et identités en exil. Ces lieux favorisent l’engagement social et professionnel tout en renforçant les liens communautaires. Cependant, leurs pratiques sont marquées par des tensions internes et des perceptions ambiguës de la part la société géorgienne. Cette étude explore les dynamiques de socialisation, les différends intercommunautaires et la visibilité de ces espaces dans le contexte politique et social local
Continuized Nesterov Acceleration for Non-Convex Optimization
No full textIn convex optimization, continuous-time counterparts have been a fruitful tool for analyzing momentum algorithms. Fewer such examples are available when the function to minimize is non-convex. In several cases, discrepancies arise between the existing discrete-time results -namely those obtained for momentum algorithms -and their continuous-time counterparts, with the latter typically yielding stronger guarantees. We argue that the continuized framework (Even et al., 2021), mixing continuous and discrete components, can tighten the gap between known continuous and discrete results. This framework relies on computations akin to standard Lyapunov analyses, from which are deduced convergence bounds for an algorithm that can be written as a Nesterov momentum algorithm with stochastic parameters. In this work, we extend the range of applicability of the continuized framework, e.g. by allowing it to handle non-smooth Lyapunov functions. We then strengthen its trajectory-wise guarantees for linear convergence rate, deriving finite time bounds with high probability and asymptotic almost sure bounds. We apply this framework to the non-convex class of strongly quasar-convex functions. Adapting continuous-time results that have weaker discrete equivalents to the continuized method, we improve by a constant factor the known convergence rate, and relax the existing assumptions on the set of minimizers
Characterization of Sub-Millimeter Subsurface Cracks in Rail Using Contrast Enhanced Ultrafast Ultrasound Imaging
No full textInternational audiencePrecise measurements of crack morphology, such as width and depth in rails, are crucial for maintaining safety and extending the service life of rail systems. Conventional nondestructive ultrasound imaging struggles to detect near-surface sub-millimeter flaws. In this study, we propose contrast-enhanced ultrafast ultrasound (CEUUS) imaging as a novel technique for nondestructive rail evaluation. Unlike conventional ultrasound methods that rely on signal responses from the defects, our approach rapidly acquires thousands of frames in seconds, enabling spatiotemporal tracking of ultrasound contrast agents (UCAs) from background noise. Localized UCAs were tracked across frames, and their trajectories were reconstructed to map the crack morphology with sub-millimeter precision. Our results demonstrate that CEUUS imaging can detect sub-millimeter crack geometries, providing insight into crack width and depth. These results highlight the potential of CEUUS to overcome longstanding challenges in detecting subsurface cracks, enabling earlier intervention and more reliable structural health monitoring in rail systems
GLUS: Guidance Loss Driven Plug-and-Play Diffusion for Ultrasound Image Super-Resolution
No full textInternational audienceThis paper introduces GLUS, a novel training-free plugand-play framework for ultrasound (US) super-resolution. GLUS extends the residual plug-and-play paradigm by incorporating a tailored guidance energy function to capture the domain-specific characteristics of US data. The proposed framework operates without the need for pre-training, offering a practical and efficient solution for enhancing US image resolution. Experiments on both simulated and in vivo data show that GLUS consistently outperforms existing benchmarks, achieving superior quantitative and qualitative results
Seeing yew for the forest: a call to action for improving conservation and restoration of the European yew (Taxus baccata L.)
No full textInternational audienceThe European yew (Taxus baccata L.) is a long-lived conifer of ecological, cultural, and historical importance across Eurasia. Despite its remarkable resilience, wide distribution, and symbolic importance, the species has experienced a long-term decline due to a complex interplay of climatic fluctuations, megafaunal extinctions, human exploitation, and insufficient regeneration. Recent studies in palaeoecology, archaeology, dendroecology, and conservation have revealed a species with greater ecological plasticity and a broader historical distribution than previously assumed. However, many fundamental questions remain unresolved, particularly regarding its biogeographical history, population dynamics, recruitment processes, and the drivers of its decline.This review stems from prior investigations of yew in the French Pyrenees and, more broadly, across Europe. These efforts led to a transdisciplinary seminar and opened a collaboration uniting >30 researchers across Eurasia. By synthesizing a wide array of data and perspectives, the article highlights key knowledge gaps and outlines emerging research priorities. These are organized thematically—past, present, and future—and include 25 questions on the species' ecological niche, life-history strategies, human interactions, genetic resilience, and conservation under global change. The article advocates for a shift towards integrative and long-term conservation strategies that embrace the historical legacies of yew populations, the general ecology of the species along with local ecological context dependence, and the urgency of future threats. By identifying pressing research needs, this review seeks to lay the foundation for new collaborative initiatives and to support evidence-based conservation of this emblematic yet understudied species
BORWin: Exact algorithm based on a Bi-Objective Relaxation for Window-constrained problems
No full textInternational audienceA mixed integer maximization problem involving several additional constraints defined with both a lower and an upper bound is considered. It is assumed that one of such constraints is more restrictive than the others. As it can be seen as a resource window constraint, it defines the so-called window-constrained problem. From a bi-objective perspective, a 2-phase algorithm, called BORWin, is devised. It stands for Bi-Objective Relaxation for Window-constrained problems. The first phase is generic for any window-constrained problem and provides a family of upper bounds based on a bi-objective relaxation of the additional constraints. It is shown that the latter bounds strongly relate to the Lagrangian dual bounds. The second phase is derived for a variant involving a graph structure, namely the window-constrained longest-path problem on an acyclic graph. The aim is to take advantage of the upper bounds to devise an efficient label extension algorithm. It is shown that complementary upper bounds could be derived to further improve performance in some special cases. A typical example is when the additional constraints have special knapsack structures. This is the case for the Hydro-Unit Commitment problem with a single plant (1-HUC). From numerical experiments for the 1-HUC, BOR-Win appears to be very efficient compared to state-of-the-art approaches
Causal Deterministic SLAM: Representation and Compensation of Independent Systems of Errors
No full textThis paper addresses the emblematic Simultaneous Localization And Mapping (SLAM) problem through the lens of causality. The proposed approach explicitly represents the causal deterministic relationships on the robot's relative motion and actuator/sensor signals, together with the effect of exogenous random errors, into a structural causal model and associated causal graph. The SLAM solution is obtained by "compensating" these errors. This adjustment process can take place only when a loop closure occurs, and proceeds in two stages: first, the errors are partitioned into conditionally independent subsets through d-separation analysis; then, each obtained subset is compensated separately by means of a dedicated algorithm which ensures the respect of Gaussian assumptions and the satisfaction of constraints. The estimation of the robot trajectory and landmark poses comes as a by-product of error compensation. Throughout the paper, the conceptual differences with the standard probabilistic view of SLAM are highlighted. The strengths and limits of the alternative causal approach are discussed, and illustrated on experiments built with synthetic and real data
AcouBLE: Clock-Seeding BLE chips via Acoustic Signals for Extended Energy Efficiency
No full textInternational audienceWe present AcouBLE, the first system to significantly reduce the BLE tag's RF front-end energy consumption while preserving standard BLE operation, backward compatibility, and one-to-one deployability. At its core is clock-seeding, a new paradigm that offloads the generation of the high-frequency MHz clock from the energy-constrained tag to the central (smartphone). The smartphone transmits a low-frequency audio signal that the tag transforms into a clean, ultra-low-jitter reference clock-eliminating the most energy-expensive component of a BLE connection event.Our design combines a novel digital-analog upconversion pipeline that converts a kHz audio tone into a stable 32 MHz clock with a lightweight feedback protocol to implement clock-seeding on commodity BLE chips without modifying the BLE stack or requiring privileged access on the central. A full prototype demonstrates up to 5.9X energy savings in Connected mode, while maintaining RF performance, seamless deployability, and backward compatibility.</p