Showing papers by "Grenoble Institute of Technology published in 2019"

Riemannian Procrustes Analysis: Transfer Learning for Brain–Computer Interfaces

Abstract: Objective: This paper presents a Transfer Learning approach for dealing with the statistical variability of electroencephalographic (EEG) signals recorded on different sessions and/or from different subjects. This is a common problem faced by brain–computer interfaces (BCI) and poses a challenge for systems that try to reuse data from previous recordings to avoid a calibration phase for new users or new sessions for the same user. Method: We propose a method based on Procrustes analysis for matching the statistical distributions of two datasets using simple geometrical transformations (translation, scaling, and rotation) over the data points. We use symmetric positive definite matrices (SPD) as statistical features for describing the EEG signals, so the geometrical operations on the data points respect the intrinsic geometry of the SPD manifold. Because of its geometry-aware nature, we call our method the Riemannian Procrustes analysis (RPA). We assess the improvement in transfer learning via RPA by performing classification tasks on simulated data and on eight publicly available BCI datasets covering three experimental paradigms (243 subjects in total). Results: Our results show that the classification accuracy with RPA is superior in comparison to other geometry-aware methods proposed in the literature. We also observe improvements in ensemble classification strategies when the statistics of the datasets are matched via RPA. Conclusion and significance: We present a simple yet powerful method for matching the statistical distributions of two datasets, thus paving the way to BCI systems capable of reusing data from previous sessions and avoid the need of a calibration procedure.

KTlO: a metal shrouded 2D semiconductor with high carrier mobility and tunable magnetism

Abstract: Two-dimensional materials with high carrier mobility and tunable magnetism are in high demand for nanoelectronic and spintronic applications. Herein, we predict a novel two-dimensional monolayer KTlO that possesses an indirect band gap of 2.25 eV (based on HSE06 calculations) and high carrier mobility (450 cm2 V−1 s−1 for electrons and 160 cm2 V−1 s−1 for holes) by means of ab initio calculations. The electron mobility can be increased up to 26 280 cm2 V−1 s−1 and 54 150 cm2 V−1 s−1 for bilayer and trilayer KTlO, respectively. The KTlO monolayer has a calculated cleavage energy of 0.56 J m−2, which suggests exfoliation of the bulk material as a viable means for the preparation of mono- and few-layer materials. Remarkably, the KTlO monolayer demonstrates tunable magnetism and half-metallicity with hole doping, which are attributed to the novel Mexican-hat-like bands and van Hove singularities in its electronic structure. Furthermore, monolayer KTlO exhibits moderate optical absorption over the visible light and ultraviolet regions. The band gap value and band characteristics of monolayer KTlO can be substantially manipulated by biaxial and uniaxial strains to meet the requirement of various applications. All these novel properties make monolayer KTlO a promising functional material for future nanoelectronic and spintronic applications.

Introduction to Distributed Self-Stabilizing Algorithms

Abstract: This book aims at being a comprehensive and pedagogical introduction to the concept of self-stabilization, introduced by Edsger Wybe Dijkstra in 1973. Self-stabilization characterizes the ...

The concept of electrostatic doping and related devices

Abstract: Electrostatic doping aims at replacing donor/acceptor dopant species with free electron/hole charges induced by the gates in ultrathin MOS structures. Highly doped N+/P+ terminals and virtual P-N junctions can be emulated in undoped layers prompting innovative reconfigurable devices with enriched functionality. The distinct merit is that the carrier concentration and polarity (i.e., electrostatic doping) are tunable via the gate bias. After presenting the fundamentals, we review the family of electrostatically-doped devices fabricated with emerging or mature technologies (nanowires, nanotubes, 2D materials, FD-SOI). The multiple facets of the Hocus Pocus diode are discussed by underlining the difference with classical physical diodes. Electrostatic doping gave rise to a number of band-modulation devices with outstanding memory and sharp-switching capability. The concept, intrinsic mechanisms and typical applications are described in detail.

Artificial skin through super-sensing method and electrical impedance data from conductive fabric with aid of deep learning

Abstract: Sense of touch is a major part of man's communication with their environment. Artificial skins can help robots to have the same sense of touch, especially for their social interactions. This paper presents a pressure mapping sensing using piezo-resistive fabric to represent aspects of the sense of touch. In past few years' electrical impedance tomography (EIT) is considered to be able offer a good alternative for artificial skin in particular for its ease of adaptation for large area skin compared to individual matrix based sensors. The EIT has also very good temporal performance in data collection allowing for monitoring of fast responses to touch stimulation, enabling a truly real time touch sensing. Electromechanical responses of a conductive fabric can be exploited using EIT to create a low cost and large area touch sensing. Such electromechanical properties are often very complex, so to improve the imaging resolution and touch visibility an artificial intelligent (AI) was used in addition to the state of the art spatio-temporal imaging algorithm. This work demonstrates a step towards an integrated seamless skin with large area sensing in dynamical settings, closer to natural human skin's behaviour. For the first time a dynamical touch sensing are studies by means of a spatio-temporal based electrical impedance tomography (EIT) imaging on a conductive fabric. The experimental results demonstrated the successful results by a combined AI with dynamical EIT imaging results in single and multiple points of touch.

Voltage Control in Low-Voltage Grids Using Distributed Photovoltaic Converters and Centralized Devices

Abstract: This paper studies the application of distributed and centralized solutions for voltage control in low voltage (LV) grids with high photovoltaic (PV) penetration. In traditional LV grids, the coordination of distributed PV converters and a centralized device would require massive investments in new communication and control infrastructures. The alternative of exploiting distributed PV converters for voltage control is discussed, showing that it can help to stabilize the voltage in the grid connection points also without coordination between them and/or with a centralized unit. The goal of this paper is to investigate how the setup of the voltage controllers inside PV inverters affects the operation of these controllers taking into account the limits for reactive power injection. In addition, the interaction of distributed PV converters with centralized devices (static var compensators and on load tap changers) is analyzed to assess whether additional benefits may come in these cases.

Sentinel-2 Sharpening Using a Reduced-Rank Method

Abstract: Recently, the Sentinel-2 (S2) satellite constellation was deployed for mapping and monitoring the Earth environment. Images acquired by the sensors mounted on the S2 platforms have three levels of spatial resolution: 10, 20, and 60 m. In many remote sensing applications, the availability of images at the highest spatial resolution (i.e., 10 m for S2) is often desirable. This can be achieved by generating a synthetic high-resolution image through data fusion. To this end, researchers have proposed techniques exploiting the spectral/spatial correlation inherent in multispectral data to sharpen the lower resolution S2 bands to 10 m. In this paper, we propose a novel method that formulates the sharpening process as a solution to an inverse problem. We develop a cyclic descent algorithm called S2Sharp and an associated tuning parameter selection algorithm based on generalized cross validation and Bayesian optimization. The tuning parameter selection method is evaluated on a simulated data set. The effectiveness of S2Sharp is assessed experimentally by comparisons to state-of-the-art methods using both simulated and real data sets.

Assessing the performances of different neural network architectures for the detection of screams and shouts in public transportation

Abstract: As intelligent transportation systems are becoming more and more prevalent, the relevance of automatic surveillance systems grows larger While such systems rely heavily on video signals, other types of signals can be used as well to monitor the security of passengers The present article proposes an audio-based intelligent system for surveillance in public transportation, investigating the use of some state-of-the-art artificial intelligence methods for the automatic detection of screams and shouts We present test results produced on a database of sounds occurring in subway trains in real working conditions, by classifying sounds into screams, shouts and other categories using different Neural Network architectures The relevance of these architectures in the analysis of audio signals is analyzed We report encouraging results, given the difficulty of the task, especially when a high level of surrounding noise is present

Review of Machine Learning Approaches In Fault Diagnosis applied to IoT Systems

Abstract: With increasing complex systems, low production costs, and changing technologies, for this reason, the automatic fault diagnosis using artificial intelligence (AI) techniques is more in more applied. In addition, with the emergence of the use of reconfigurable systems, AI can assist in self-maintenance of complex systems. The purpose of this article is to summarize the diagnosis research of systems using AI approaches and examine their application particularly in the field of diagnosis of complex systems. It covers articles published from 2002 to 2018 using Machine Learning tools for fault diagnosis in industrial systems.

Implementing an Impedance Compression Network to Compensate for Misalignments in a Wireless Power Transfer System

Abstract: This paper presents an impedance compression network (ICN) design in a wireless power transfer (WPT) system to compensate for distance or alignment variations between coils. In midrange WPT applications, magnetic resonant coupling coils are mainly implemented to achieve high efficiency for charging batteries. However, a distance or horizontal alignment variation between coils changes their coupling coefficient, and it decreases the overall performance of WPT systems because the zero voltage switching in a resonant inverter is lost. In order to reduce coil impedance variation, we propose an ICN that compresses variations in coil impedance. The ICN consists of a resistance compression network and a phase compression network to suppress magnitude and phase variations, respectively, in the coil impedance. Only passive components, such as inductors and capacitors, are used to implement an ICN.

The Riemannian Potato Field: A Tool for Online Signal Quality Index of EEG

Abstract: Electroencephalographic (EEG) recordings are contaminated by instrumental, environmental, and biological artifacts, resulting in low signal-to-noise ratio. Artifact detection is a critical task for real-time applications where the signal is used to give a continuous feedback to the user. In these applications, it is therefore necessary to estimate online a signal quality index (SQI) in order to stop the feedback when the signal quality is unacceptable. In this paper, we introduce the Riemannian potato field (RPF) algorithm as such SQI. It is a generalization and extensionof theRiemannian potato, a previouslypublished real-time artifact detection algorithm, whose performance is degraded as the number of channels increases. The RPF overcomes this limitation by combining the outputs of several smaller potatoes into a unique SQI resulting in a higher sensitivity and specificity, regardless of the number of electrodes. We demonstrate these results on a clinical dataset totalizing more than 2200 h of EEG recorded at home, that is, in a non-controlled environment.

Semi-Automatic Annotation of Objects in Visual-Thermal Video

Abstract: Deep learning requires large amounts of annotated data. Manual annotation of objects in video is, regardless of annotation type, a tedious and time-consuming process. In particular, for scarcely used image modalities human annotation is hard to justify. In such cases, semi-automatic annotation provides an acceptable option. In this work, a recursive, semi-automatic annotation method for video is presented. The proposed method utilizes a state-of-the-art video object segmentation method to propose initial annotations for all frames in a video based on only a few manual object segmentations. In the case of a multi-modal dataset, the multi-modality is exploited to refine the proposed annotations even further. The final tentative annotations are presented to the user for manual correction. The method is evaluated on a subset of the RGBT-234 visual-thermal dataset reducing the workload for a human annotator with approximately 78% compared to full manual annotation. Utilizing the proposed pipeline, sequences are annotated for the VOT-RGBT 2019 challenge.

Planar penta-transition metal phosphide and arsenide as narrow-gap semiconductors with ultrahigh carrier mobility

Abstract: Searching for single-atom thin materials in the planar structure, like graphene and borophene, is one of the most attractive themes in two-dimensional materials. Using density functional theory calculations, we have proposed a series of single-layer planar penta-transition metal phosphides and arsenides, i.e., TM2X4 (TM = Ni, Pd and Pt; X = P, As). According to the calculated phonon dispersion relation and elastic constants, as well as ab initio molecular dynamics simulation results, monolayers of planar penta-TM2X4 are dynamically, mechanically and thermally stable. In addition, screened HSE06 hybrid functional calculations including spin–orbit coupling show that these monolayers are direct band gap semiconductors, with band gaps ranging from 0.14 to 0.77 eV. Ultrahigh carrier mobilities up to 104–105 cm2 V−1 s−1 both for electrons and holes have been confirmed, among the highest in 2D semiconductors. Our results indicate that planar penta-TM2X4 monolayers are interesting narrow-gap semiconductors with ultrahigh carrier mobility as well as excellent optical properties, and may find potential applications in nanoelectronics and photoelectronics.

Robustness to In-Domain Viscous Damping of a Collocated Boundary Adaptive Feedback Law for an Antidamped Boundary Wave PDE

Abstract: In this paper, the robustness to model mismatch of a preexisting collocated boundary adaptive feedback law is investigated. This control law was originally designed for an antidamped pure wave partial differential equation (PDE). Actuation and measurements are located at the same boundary. Adaptive terms account for uncertain parameters located at the antidamped boundary, opposite to the collocated actuation and measurement. By extending and transforming the system state using, in particular, backstepping, this paper establishes that this controller is robust to sufficiently small in-domain damping. In particular, both stability and attractivity (convergence) are established similarly as in the nominal case. Note moreover that, assuming that some parameters are known, the exponential stability to an attractor holds. Simulations are performed to illustrate the interest of this study to attenuate mechanical vibrations in an oil-drilling context.

Design of experiments to assess the effect of culture parameters on the osteogenic differentiation of human adipose stromal cells.

Abstract: Background Human adipose-derived stromal cells (hASCs) have been gaining increasing popularity in regenerative medicine thanks to their multipotency, ease of collection, and efficient culture. Similarly to other stromal cells, their function is particularly sensitive to the culture conditions, including the composition of the culture medium. Given the large number of parameters that can play a role in their specification, the rapid assessment would be beneficial to allow the optimization of their culture parameters.

Fast and exact analysis for LRU caches

Abstract: For applications in worst-case execution time analysis and in security, it is desirable to statically classify memory accesses into those that result in cache hits, and those that result in cache misses. Among cache replacement policies, the least recently used (LRU) policy has been studied the most and is considered to be the most predictable. The state-of-the-art in LRU cache analysis presents a tradeoff between precision and analysis efficiency: The classical approach to analyzing programs running on LRU caches, an abstract interpretation based on a range abstraction, is very fast but can be imprecise. An exact analysis was recently presented, but, as a last resort, it calls a model checker, which is expensive. In this paper, we develop an analysis based on abstract interpretation that comes close to the efficiency of the classical approach, while achieving exact classification of all memory accesses as the model-checking approach. Compared with the model-checking approach we observe speedups of several orders of magnitude. As a secondary contribution we show that LRU cache analysis problems are in general NP-complete.

Dual-output DC/DC boost converter for bipolar DC microgrids

Abstract: This study presents a new step-up DC/DC converter for bipolar DC microgrids. This relies on utilising a new converter topology which offers bipolar DC outputs with the key features of boosting the input voltage and balancing the output voltages. In addition, the proposed converter can transfer power from output terminals to the input side which leads to energy saving. Due to the microgrid's loading conditions, different modes can be considered, and the operational principles are analysed in details. Applying the circuit average method, the small signal model of the proposed converter is derived. Then the appropriate controllers are designed considering the small signal model of the converter in each mode. Finally, a prototype of the proposed converter has been implemented in the laboratory, and the experimental results have verified the converter's ability to balance the bipolar DC grid's voltages and its capability of transferring power in various microgrid's conditions.

Intrinsic Cramér–Rao Bounds for Scatter and Shape Matrices Estimation in CES Distributions

Abstract: Scatter matrix and its normalized counterpart, referred to as shape matrix, are key parameters in multivariate statistical signal processing, as they generalize the concept of covariance matrix in the widely used Complex Elliptically Symmetric distributions. Following the framework of [1] , intrinsic Cramer–Rao bounds are derived for the problem of scatter and shape matrices estimation with samples following a Complex Elliptically Symmetric distribution. The Fisher Information Metric and its associated Riemannian distance (namely, CES-Fisher) on the manifold of Hermitian positive definite matrices are derived. Based on these results, intrinsic Cramer–Rao bounds on the considered problems are then expressed for three different distances (Euclidean, natural Riemannian, and CES-Fisher). These contributions are therefore a generalization of Theorems 4 and 5 of [1] to a wider class of distributions and metrics for both scatter and shape matrices.

Grid Frequency Support From V2G Aggregators and HVdc Links in Presence of Nonsynchronous Units

Abstract: Massive penetration of nonsynchronous units with power electronics devices in the future grid could diminish the rotational inertia of a power system and inevitably jeopardize the frequency stability. Several adaptations are made accordingly in the power system to provide a short-term inertial support. In this paper, synchronous power concept is used for inertia emulation (IE) in high voltage direct current (HVdc) transmission links for frequency control. However, the energy reserves capacity such as direct current (dc) link capacitances or distributed generations for IE are limited and therefore required additional energy storage. This paper considered mobile electric vehicles (EVs) network as an energy storage and their charging/discharging is done by balance state-of-charge (SOC) holder technique taking into consideration future driving demand of EV owners using battery SOC controller. Cooperative contributions from HVdc links and EVs during perturbation exceptionally damp the oscillation and peak deviation in grid frequency as well as in tie-line power. To validate the efficacy of proposed framework, numerous time domain simulation results are presented.

Estimating Occupancy Using Interactive Learning With a Sensor Environment: Real-Time Experiments

Abstract: Interactive learning plays a key role in extending the occupant behavior implementation toward smart buildings. Efficient feedbacks can be obtained from the end user by involving occupants and increasing their awareness about energy systems. Working in highly energy-efficient buildings can be a great opportunity, but users need to feel empowered. This means making them aware of the building features and allowing them to manage some of the appliances. In this way, disorientation or annoyance is avoided, and people feel more in control. This paper proposes a solution to interact with occupants to estimate the number of occupants. A novel way of supervised learning is analyzed to estimate the occupancy in a room where actual occupancy is interactively requested to occupants when it is the most relevant to limit the number of interactions. Occupancy estimation algorithm relies on machine learning; it uses information gathered from occupants with the measurements collected from common sensors, for instance, motion detection, power consumption, and CO2 concentration. Two different classifiers are investigated for occupancy estimation with interactions: a decision tree C4.5 and a parameterized rule-based classifier. In this paper, the question of when interacting with occupants is investigated. This approach avoids the usage of a camera to determine the actual occupancy. A complete real-time interaction environment has been developed and is used to estimate the occupancy in an office case study. The graphical user interface has been designed to carry out a real-time experiment.

Wave Boundary Layer Hydrodynamics and Sheet Flow Properties under Large‐Scale Plunging‐Type Breaking Waves

Abstract: Wave boundary layer (WBL) dynamics are measured with an Acoustic Concentration and Velocity Profiler (ACVP) across the sheet flow-dominated wave-breaking region of regular large-scale waves breaking as a plunger over a developing breaker bar. Acoustic sheet flow measurements are first evaluated quantitatively in comparison to Conductivity Concentration Meter (CCM+) data used as a reference. The near-bed orbital velocity field exhibits expected behaviors in terms of wave shape, intrawave WBL thickness, and velocity phase leads. The observed fully turbulent flow regime all across the studied wave-breaking region supports the model-predicted transformation of free-stream velocity asymmetry into near-bed velocity skewness inside the WBL. Intrawave concentration dynamics reveal the existence of a lower pickup layer and an upper sheet flow layer similar to skewed oscillatory sheet flows, and with similar characteristics in terms of erosion depth and sheet flow layer thickness. Compared to the shoaling region, differences in terms of sheet flow and hydrodynamic properties of the flow are observed at the plunge point, attributed to the locally enhanced wave breaker turbulence. The ACVP-measured total sheet flow transport rate is decomposed into its current-, wave-, and turbulence-driven components. In the shoaling region, the sand transport is found to be fully dominated by the onshore skewed wave-driven component with negligible phase lag effects. In the outer surf zone, the total net flux exhibits a three-layer vertical structure typical of skewed oscillatory sheet flows. However, in the present experiments this structure originates from offshore-directed undertow-driven flux, rather than from phase lag effects.