École Centrale de Lille

We introduce a general-purpose conditioning method for neural networks called FiLM: Feature-wise Linear Modulation. FiLM layers influence neural network computation via a simple, feature-wise affine transformation based on conditioning information. We show that FiLM layers are highly effective for visual reasoning - answering image-related questions which require a multi-step, high-level process - a task which has proven difficult for standard deep learning methods that do not explicitly model reasoning. Specifically, we show on visual reasoning tasks that FiLM layers 1) halve state-of-the-art error for the CLEVR benchmark, 2) modulate features in a coherent manner, 3) are robust to ablations and architectural modifications, and 4) generalize well to challenging, new data from few examples or even zero-shot.

The development of energy management tools for next-generation PhotoVoltaic (PV) installations, including storage units, provides flexibility to distribution system operators. In this paper, the aggregation and implementation of these determinist energy management methods for business customers in a microgrid power system are presented. This paper proposes a determinist energy management system for a microgrid, including advanced PV generators with embedded storage units and a gas microturbine. The system is organized according to different functions and is implemented in two parts: a central energy management of the microgrid and a local power management at the customer side. The power planning is designed according to the prediction for PV power production and the load forecasting. The central and local management systems exchange data and order through a communication network. According to received grid power references, additional functions are also designed to manage locally the power flows between the various sources. Application to the case of a hybrid supercapacitor battery-based PV active generator is presented.

Furfural and 5-hydroxymethylfurfural stand out as bridges connecting biomass raw materials to the biorefinery industry. Their reductive transformations by hydroconversion are key routes toward a wide variety of chemicals and biofuels, and heterogeneous catalysis plays a central role in these reactions. The catalyst efficiency highly depends on the nature of metals, supports, and additives, on the catalyst preparation procedure, and obviously on reaction conditions to which catalyst and reactants are exposed: solvent, pressure, and temperature. The present review focuses on the roles played by the catalyst at the molecular level in the hydroconversion of furfural and 5-hydroxymethylfurfural in the gas or liquid phases, including catalytic hydrogen transfer routes and electro/photoreduction, into oxygenates or hydrocarbons (e.g., furfuryl alcohol, 2,5-bis(hydroxymethyl)furan, cyclopentanone, 1,5-pentanediol, 2-methylfuran, 2,5-dimethylfuran, furan, furfuryl ethers, etc.). The mechanism of adsorption of the reactant and the mechanism of the reaction of hydroconversion are correlated to the specificities of each active metal, both noble (Pt, Pd, Ru, Au, Rh, and Ir) and non-noble (Ni, Cu, Co, Mo, and Fe), with an emphasis on the role of the support and of additives on catalytic performances (conversion, yield, and stability). The reusability of catalytic systems (deactivation mechanism, protection, and regeneration methods) is also discussed.

This paper presents a method for detection of steganographic methods that embed in the spatial domain by adding a low-amplitude independent stego signal, an example of which is least significant bit (LSB) matching. First, arguments are provided for modeling the differences between adjacent pixels using first-order and second-order Markov chains. Subsets of sample transition probability matrices are then used as features for a steganalyzer implemented by support vector machines. The major part of experiments, performed on four diverse image databases, focuses on evaluation of detection of LSB matching. The comparison to prior art reveals that the presented feature set offers superior accuracy in detecting LSB matching. Even though the feature set was developed specifically for spatial domain steganalysis, by constructing steganalyzers for ten algorithms for JPEG images, it is demonstrated that the features detect steganography in the transform domain as well.

Traditionally, assignment and scheduling decisions are made separately at different levels of the production management framework. The combining of such decisions presents additional complexity and new problems. We present two new approaches to solve jointly the assignment and job-shop scheduling problems (with total or partial flexibility). The first one is the approach by localization (AL). It makes it possible to solve the problem of resource allocation and build an ideal assignment model (assignments schemata). The second one is an evolutionary approach controlled by the assignment model (generated by the first approach). In such an approach, we apply advanced genetic manipulations in order to enhance the solution quality. We also explain some of the practical and theoretical considerations in the construction of a more robust encoding that will enable us to solve the flexible job-shop problem by applying the genetic algorithms (GAs). Two examples are presented to show the efficiency of the two suggested methodologies.

), a calcium phosphate biomaterial, is a very promising candidate for the treatment of air, water and soil pollution. Indeed, hydroxyapatite (Hap) can be extremely useful in the field of environmental management, due in one part to its particular structure and attractive properties, such as its great adsorption capacities, its acid-base adjustability, its ion-exchange capability and its good thermal stability. Moreover, Hap is able to constitute a valuable resource recovery route. The first part of this review will be dedicated towards presenting Hap's structure and defining properties that result in its viability as an environmental remediation material. The second will focus on its use as adsorbent for wastewater and soil treatment, while indicating the mechanisms involved in this remediation process. Finally, the last part will impart all findings on Hap's applications in the field of catalysis, whether it be as catalyst, as photocatalyst, or as active phase support. Hence, all of the above will have served in showcasing the benefits gained by employing hydroxyapatite in air, water and soil clean-up.

In this paper, we study the dynamics of cylindrical armoured bubbles excited by mechanical vibrations. A step by step transition from cylindrical to spherical shape is reported as the intensity of the vibration is increased, leading to a reduction of the bubble surface and a dissemination of the excess particles. We demonstrate through energy balance that nonspherical armoured bubbles constitute a metastable state. The vibration instills the activation energy necessary for the bubble to return to its least energetic stable state: a spherical armoured bubble. At this point, particle desorption can only be achieved through higher amplitude of excitation required to overcome capillary retention forces. Nonspherical armoured bubbles open perspectives for tailored localized particle dissemination with limited excitation power.

We introduce GuessWhat?!, a two-player guessing game as a testbed for research on the interplay of computer vision and dialogue systems. The goal of the game is to locate an unknown object in a rich image scene by asking a sequence of questions. Higher-level image understanding, like spatial reasoning and language grounding, is required to solve the proposed task. Our key contribution is the collection of a large-scale dataset consisting of 150K human-played games with a total of 800K visual question-answer pairs on 66K images. We explain our design decisions in collecting the dataset and introduce the oracle and questioner tasks that are associated with the two players of the game. We prototyped deep learning models to establish initial baselines of the introduced tasks.

Over the last few years, extraordinary advances in experimental and theoretical tools have allowed us to monitor and control matter at short time and atomic scales with a high degree of precision. An appealing and challenging route toward engineering materials with tailored properties is to find ways to design or selectively manipulate materials, especially at the quantum level. To this end, having a state-of-the-art ab initio computer simulation tool that enables a reliable and accurate simulation of light-induced changes in the physical and chemical properties of complex systems is of utmost importance. The first principles real-space-based Octopus project was born with that idea in mind, i.e., to provide a unique framework that allows us to describe non-equilibrium phenomena in molecular complexes, low dimensional materials, and extended systems by accounting for electronic, ionic, and photon quantum mechanical effects within a generalized time-dependent density functional theory. This article aims to present the new features that have been implemented over the last few years, including technical developments related to performance and massive parallelism. We also describe the major theoretical developments to address ultrafast light-driven processes, such as the new theoretical framework of quantum electrodynamics density-functional formalism for the description of novel light-matter hybrid states. Those advances, and others being released soon as part of the Octopus package, will allow the scientific community to simulate and characterize spatial and time-resolved spectroscopies, ultrafast phenomena in molecules and materials, and new emergent states of matter (quantum electrodynamical-materials).

Fischer-Tropsch synthesis (FTS) is an essential approach to convert coal, biomass, and shale gas into fuels and chemicals, such as lower olefins, gasoline, diesel, and so on. In recent years, there has been increasing motivation to deploy FTS at commercial scales which has been boosting the discovery of high performance catalysts. In particular, the importance of support in modulating the activity of metals has been recognized and carbonaceous materials have attracted attention as supports for FTS. In this review, we summarised the substantial progress in the preparation of carbon-based catalysts for FTS by applying activated carbon (AC), carbon nanotubes (CNTs), carbon nanofibers (CNFs), carbon spheres (CSs), and metal-organic frameworks (MOFs) derived carbonaceous materials as supports. A general assessment of carbon-based catalysts for FTS, concerning the support and metal properties, activity and products selectivity, and their interactions is systematically discussed. Finally, current challenges and future trends in the development of carbon-based catalysts for commercial utilization in FTS are proposed.

The fabrication of miniaturized electrochemical energy storage systems is essential for the development of future electronic devices for Internet of Thing applications. This paper aims at reviewing the current micro-supercapacitor technologies and at defining the guidelines to produce high performance micro-devices with special focuses onto the 3D designs as well as the fabrication of solid state miniaturized devices to solve the packaging issue.

In this paper, we study the dynamics of cylindrical armoured bubbles excited by mechanical vibrations. A step by step transition from cylindrical to spherical shape is reported as the intensity of the vibration is increased, leading to a reduction of the bubble surface and a dissemination of the excess particles. We demonstrate through energy balance that nonspherical armoured bubbles constitute a metastable state. The vibration instills the activation energy necessary for the bubble to return to its least energetic stable state: a spherical armoured bubble. At this point, particle desorption can only be achieved through higher amplitude of excitation required to overcome capillary retention forces. Nonspherical armoured bubbles open perspectives for tailored localized particle dissemination with limited excitation power.

Presently, conventional technologies in water treatment are not efficient enough to completely mineralize refractory water contaminants. In this context, the implementation of catalytic processes could be an alternative. Despite the advantages provided in terms of kinetics of transformation, selectivity, and energy saving, numerous attempts have not yet led to implementation at an industrial scale. This review examines investigations at different scales for which controversies and limitations must be solved to bridge the gap between fundamentals and practical developments. Particular attention has been paid to the development of solar-driven catalytic technologies and some other emerging processes, such as microwave assisted catalysis, plasma-catalytic processes, or biocatalytic remediation, taking into account their specific advantages and the drawbacks. Challenges for which a better understanding related to the complexity of the systems and the coexistence of various solid-liquid-gas interfaces have been identified.

This paper considers the control of a group of autonomous mobile robots. A coordinated control scheme based on a leader-follower approach is developed to achieve formation maneuvers. First and second order sliding-mode controllers are proposed for asymptotically stabilizing the vehicles to a time-varying desired formation. The latter controller, based on the relative motion states, eliminates the need for measurement or estimation of the leader velocity. It enables formation stabilization using a vision system carried by the followers and ensures the collision avoidance from the initial time instance. Experimental investigation has been conducted using a test bench made of three nonholonomic mobile robots in order to demonstrate the effectiveness of the proposed strategy.

Classical wind energy conversion systems are usually passive generators. The generated power does not depend on the grid requirement but entirely on the fluctuant wind condition. A dc-coupled wind/hydrogen/supercapacitor hybrid power system is studied in this paper. The purpose of the control system is to coordinate these different sources, particularly their power exchange, in order to make controllable the generated power. As a result, an active wind generator can be built to provide some ancillary services to the grid. The control system should be adapted to integrate the power management strategies. Two power management strategies are presented and compared experimentally. We found that the “source-following” strategy has better performances on the grid power regulation than the “grid-following” strategy.

. The underlying mechanism of action of these CQDs was revealed to be inhibition of HCoV-229E entry that could be due to interaction of the functional groups of the CQDs with HCoV-229E entry receptors; surprisingly, an equally large inhibition activity was observed at the viral replication step.

Asymptotic stability of a class of linear equations with arbitrary discrete and distributed delays is investigated. Both delay-independent and delay-dependent stability conditions are formulated in terms of existence of positive definite solutions to Riccati matrix equations. The approach of deriving various Riccati equations using the direct Lyapunov method is proposed.

The problem of asymptotic tracking control for a class of uncertain switched nonlinear systems under fuzzy approximation framework is solved in this paper. Superior to most existing results based on fuzzy adaptive control strategy that can only achieve bounded error tracking performance, our proposed control scheme can guarantee the local asymptotic tracking performance for the uncertain switched nonlinear systems under consideration. This is accomplished by constructing a nonsmooth Lyapunov function and introducing a novel discontinuous controller with dynamic feedback compensator in the design procedure. Meanwhile, some concepts, such as differential inclusion and set-valued map, are introduced to theoretically verify the local asymptotic tracking performance of the systems with our proposed controller. With the help of set-valued Lie derivative, the common virtual control functions, the desired controller, and the adaptive laws can be precisely constructed. Finally, simulation results are given to show the effectiveness of the proposed method.

A hybrid generator with a photovoltaic energy conversion system is proposed with supercapacitors and lead-acid batteries in a dc-coupled structure. The objective of this system is to supply the prescribed reactive and active power to the grid. This paper focuses on the strategy, which makes it possible to ensure a high battery state of charge and overcharge security by designing a dedicated local control system. A continuous dynamic model and a control design of the power system studied are proposed in this paper. Simulation and experimental results illustrate the performances obtained.

PURPOSE: Little is known about injury prevention practices in professional football clubs. The purpose of this study was therefore to determine the current perceptions and practices of premier league football clubs internationally concerning risk factors, testing and preventative exercises for non-contact injuries. METHODS: A survey was administered to 93 premier league football clubs internationally. The survey included four sections: (1) persons involved in the injury prevention programme: position, quantity, role, qualification; (2) perceptions regarding non-contact injury risk factors; (3) tests used to identify non-contact injury risk and (4) non-contact injury prevention exercises used, their perceived effectiveness and implementation strategies. RESULTS: 44 surveys were successfully returned (47%). The position of physiotherapist was the most represented position in the injury prevention programme. The top five perceived risk factors in rank order were previous injury, fatigue, muscle imbalance, fitness and movement efficiency. The five most commonly used tests to identify injury risk (in rank order) were functional movement screen, questionnaire, isokinetic dynamometry, physical tests and flexibility. The top five exercises used by clubs were (also in rank order) eccentric exercise, balance/proprioception, hamstring eccentric, core stability and, sharing the fifth position, Nordic hamstring and gluteus activation. CONCLUSIONS: The survey revealed the most common perceptions and practices of premier league football clubs internationally regarding risk factors, testing and preventative exercises. The findings can enable reduction of the gap between research and practice.