Laboratoire d'Informatique, Signaux et Systèmes de Sophia Antipolis

OBJECTIVE: Most current electroencephalography (EEG)-based brain-computer interfaces (BCIs) are based on machine learning algorithms. There is a large diversity of classifier types that are used in this field, as described in our 2007 review paper. Now, approximately ten years after this review publication, many new algorithms have been developed and tested to classify EEG signals in BCIs. The time is therefore ripe for an updated review of EEG classification algorithms for BCIs. APPROACH: We surveyed the BCI and machine learning literature from 2007 to 2017 to identify the new classification approaches that have been investigated to design BCIs. We synthesize these studies in order to present such algorithms, to report how they were used for BCIs, what were the outcomes, and to identify their pros and cons. MAIN RESULTS: We found that the recently designed classification algorithms for EEG-based BCIs can be divided into four main categories: adaptive classifiers, matrix and tensor classifiers, transfer learning and deep learning, plus a few other miscellaneous classifiers. Among these, adaptive classifiers were demonstrated to be generally superior to static ones, even with unsupervised adaptation. Transfer learning can also prove useful although the benefits of transfer learning remain unpredictable. Riemannian geometry-based methods have reached state-of-the-art performances on multiple BCI problems and deserve to be explored more thoroughly, along with tensor-based methods. Shrinkage linear discriminant analysis and random forests also appear particularly useful for small training samples settings. On the other hand, deep learning methods have not yet shown convincing improvement over state-of-the-art BCI methods. SIGNIFICANCE: This paper provides a comprehensive overview of the modern classification algorithms used in EEG-based BCIs, presents the principles of these methods and guidelines on when and how to use them. It also identifies a number of challenges to further advance EEG classification in BCI.

This paper considers the problem of obtaining good attitude estimates from measurements obtained from typical low cost inertial measurement units. The outputs of such systems are characterized by high noise levels and time varying additive biases. We formulate the filtering problem as deterministic observer kinematics posed directly on the special orthogonal group SO (3) driven by reconstructed attitude and angular velocity measurements. Lyapunov analysis results for the proposed observers are derived that ensure almost global stability of the observer error. The approach taken leads to an observer that we term the direct complementary filter. By exploiting the geometry of the special orthogonal group a related observer, termed the passive complementary filter, is derived that decouples the gyro measurements from the reconstructed attitude in the observer inputs. Both the direct and passive filters can be extended to estimate gyro bias online. The passive filter is further developed to provide a formulation in terms of the measurement error that avoids any algebraic reconstruction of the attitude. This leads to an observer on SO(3), termed the explicit complementary filter, that requires only accelerometer and gyro outputs; is suitable for implementation on embedded hardware; and provides good attitude estimates as well as estimating the gyro biases online. The performance of the observers are demonstrated with a set of experiments performed on a robotic test-bed and a radio controlled unmanned aerial vehicle.

ISBN = 978-0-12-374726-6 http://www.elsevier.com/wps/find/bookdescription.cws_home/717222/description

Many image processing problems are ill-posed and must be regularized. Usually, a roughness penalty is imposed on the solution. The difficulty is to avoid the smoothing of edges, which are very important attributes of the image. In this paper, we first give conditions for the design of such an edge-preserving regularization. Under these conditions, we show that it is possible to introduce an auxiliary variable whose role is twofold. First, it marks the discontinuities and ensures their preservation from smoothing. Second, it makes the criterion half-quadratic. The optimization is then easier. We propose a deterministic strategy, based on alternate minimizations on the image and the auxiliary variable. This leads to the definition of an original reconstruction algorithm, called ARTUR. Some theoretical properties of ARTUR are discussed. Experimental results illustrate the behavior of the algorithm. These results are shown in the field of 2D single photon emission tomography, but this method can be applied in a large number of applications in image processing.

Qualinet White Paper on Definitions of Quality of Experience Output from the fifth Qualinet meeting, Novi Sad, March 12, 2013

This paper considers the question of using a nonlinear complementary filter for attitude estimation of fixed-wing unmanned aerial vehicle (UAV) given only measurements from a low-cost inertial measurement unit. A nonlinear complementary filter is proposed that combines accelerometer output for low frequency attitude estimation with integrated gyrometer output for high frequency estimation. The raw accelerometer output includes a component corresponding to airframe acceleration, occurring primarily when the aircraft turns, as well as the gravitational acceleration that is required for the filter. The airframe acceleration is estimated using a simple centripetal force model (based on additional airspeed measurements), augmented by a first order dynamic model for angle-of-attack, and used to obtain estimates of the gravitational direction independent of the airplane manoeuvres. Experimental results are provided on a real-world data set and the performance of the filter is evaluated against the output from a full GPS/INS that was available for the data set.

Abstract This work was originally motivated by a classification of tensors proposed by Richard Harshman. In particular, we focus on simple and multiple ‘bottlenecks’, and on ‘swamps’. Existing theoretical results are surveyed, some numerical algorithms are described in details, and their numerical complexity is calculated. In particular, the interest in using the enhanced line search (ELS) enhancement in these algorithms is discussed. Computer simulations feed this discussion. Copyright © 2009 John Wiley & Sons, Ltd.

Colonoscopy is the gold standard for colon cancer screening though some polyps are still missed, thus preventing early disease detection and treatment. Several computational systems have been proposed to assist polyp detection during colonoscopy but so far without consistent evaluation. The lack of publicly available annotated databases has made it difficult to compare methods and to assess if they achieve performance levels acceptable for clinical use. The Automatic Polyp Detection sub-challenge, conducted as part of the Endoscopic Vision Challenge (http://endovis.grand-challenge.org) at the international conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) in 2015, was an effort to address this need. In this paper, we report the results of this comparative evaluation of polyp detection methods, as well as describe additional experiments to further explore differences between methods. We define performance metrics and provide evaluation databases that allow comparison of multiple methodologies. Results show that convolutional neural networks are the state of the art. Nevertheless, it is also demonstrated that combining different methodologies can lead to an improved overall performance.

Empirical studies have analyzed the relationship between destination image and other evaluative constructs such as perceived value and service quality. However, the role and influence of destination image on future behavioral intentions have been of lesser interest. Consequently, this article examines the relationship between destination image, satisfaction, and future behavioral intentions of visitors to the island of Mauritius. A theoretical model is tested on a sample of 705 international visitors using structural equations modeling. The results show that destination image has a direct and an indirect influence over future behavior. Satisfaction and overall image play a mediating role between destination image and future behavior. As a result, important theoretical and practical implications for destination image research and marketing are offered.

The knowledge about fault mode behavior of an induction motor drive system is extremely important from the standpoint of improved system design, protection, and fault-tolerant control. This paper addresses the application of motor current spectral analysis for the detection and localization of abnormal electrical and mechanical conditions that indicate, or may lead to, a failure of induction motors. Intensive research effort has been for some time focused on the motor current signature analysis. This technique utilizes the results of spectral analysis of the stator current. Reliable interpretation of the spectra is difficult since distortions of the current waveform caused by the abnormalities in the induction motor are usually minute. This paper takes the initial step to investigate the efficiency of current monitoring for diagnostic purposes. The effects of stator current spectrum are described and the related frequencies determined. In the present investigation, the frequency signature of some asymmetrical motor faults are well identified using advanced signal processing techniques, such as high-resolution spectral analysis. This technique leads to a better interpretation of the motor current spectra. In fact, experimental results clearly illustrate that stator current high-resolution spectral analysis is very sensitive to induction motor faults modifying main spectral components, such as voltage unbalance and single-phasing effects.

This article is an introduction to feedback control design for a family of robotic aerial vehicles with vertical take-off and landing (VTOL) capabilities such as quadrotors, ducted-fan tail-sitters, and helicopters. Potential applications for such devices, like surveillance, monitoring, or mapping, are varied and numerous. For these applications to emerge, motion control algorithms that guarantee a good amount of robustness against state measurement/ estimation errors and unmodeled dynamics like, for example, aerodynamic perturbations, are needed. The feedback control methods considered here range from basic linear control schemes to more elaborate nonlinear control solutions. The modeling of the dynamics of these systems is first recalled and discussed. Then several control algorithms are presented and commented upon in relation to implementation issues and various operating modes encountered in practice, from teleoperated to fully autonomous flight. Particular attention is paid to the incorporation of integral-like control actions, often overlooked in nonlinear control studies despite their practical importance to render the control performance more robust with respect to unmodeled or poorly estimated additive perturbations.

This paper provides a broad perspective and analysis of the work done in control of hybrid and convertible unmanned aerial vehicles (UAVs) for the main existing designs. These flying machines are capable of vertical take off and landing (VTOL) in helicopter mode and able to transition to high-speed forward flight in airplane mode and vice versa. This paper aims at helping engineers and researchers develop flight control systems for VTOL UAVs. To this end, a historical perspective first shows the technological advances in VTOL aircraft over the years. The main VTOL concepts and state-of-art flight control methods for VTOL UAVs are presented and discussed. This study shows both the common parts and the fundamental differences in the modeling, guidance, control, and control allocation for each hybrid-VTOL-UAV type. The open challenges and the current trends in the field are highlighted. These are namely: 1) augmenting or replacing classical controllers with data-driven methods such as neural networks and machine-learning-based controllers; 2) incorporating as much knowledge of the vehicle as possible into the flight controller, for example through model predictive control or model-based nonlinear controllers; 3) a trend towards finding a unified-control approach valid in all flight modes without the need to switch among flight controllers or to perform predefined-gain scheduling, and 4) the need to mitigate control complexity and available computing resources.

A control approach is proposed for a class of underactuated vehicles in order to stabilize reference trajectories either in thrust direction, velocity, or position. The basic modeling assumption is that the vehicle is pro-pulsed via a thrust force along a single body-fixed direction and that it has full torque actuation for attitude control (i.e., a typical actuation structure for aircrafts, vertical take-off and landing (VTOL) vehicles, submarines, etc.). Additional assumptions on the external forces applied to the vehicle are also introduced for the sake of control design and stability analyses. They are best satisfied for vehicles which are subjected to an external force field (e.g., gravity) and whose shape induces lift forces with limited amplitude, unlike airplanes but as in the case of many VTOL drones. The interactions of the vehicle with the surrounding fluid are often difficult to model precisely whereas they may significantly influence and perturb its motion. By using a standard Lyapunov-based approach, novel nonlinear feedback control laws are proposed to compensate for modeling errors and perform robustly against such perturbations. Simulation results illustrating these properties on a realistic model of a VTOL drone subjected to wind gusts are reported.

A class of estimators of the Rényi and Tsallis entropies of an unknown distribution f in ℝm is presented. These estimators are based on the kth nearest-neighbor distances computed from a sample of N i.i.d. vectors with distribution f. We show that entropies of any order q, including Shannon’s entropy, can be estimated consistently with minimal assumptions on f. Moreover, we show that it is straightforward to extend the nearest-neighbor method to estimate the statistical distance between two distributions using one i.i.d. sample from each.

Independent component analysis (ICA) aims at decomposing an observed random vector into statistically independent variables. Deflation-based implementations, such as the popular one-unit FastICA algorithm and its variants, extract the independent components one after another. A novel method for deflationary ICA, referred to as RobustICA, is put forward in this paper. This simple technique consists of performing exact line search optimization of the kurtosis contrast function. The step size leading to the global maximum of the contrast along the search direction is found among the roots of a fourth-degree polynomial. This polynomial rooting can be performed algebraically, and thus at low cost, at each iteration. Among other practical benefits, RobustICA can avoid prewhitening and deals with real- and complex-valued mixtures of possibly noncircular sources alike. The absence of prewhitening improves asymptotic performance. The algorithm is robust to local extrema and shows a very high convergence speed in terms of the computational cost required to reach a given source extraction quality, particularly for short data records. These features are demonstrated by a comparative numerical analysis on synthetic data. RobustICA's capabilities in processing real-world data involving noncircular complex strongly super-Gaussian sources are illustrated by the biomedical problem of atrial activity (AA) extraction in atrial fibrillation (AF) electrocardiograms (ECGs), where it outperforms an alternative ICA-based technique.

We consider the problem of segmenting an image through the minimization of an energy criterion involving region and boundary functionals. We show that one can go from one class to the other by solving Poisson's or Helmholtz's equation with well-chosen boundary conditions. Using this equivalence, we study the case of a large class of region functionals by standard methods of the calculus of variations and derive the corresponding Euler--Lagrange equations. We revisit this problem using the notion of a shape derivative and show that the same equations can be elegantly derived without going through the unnatural step of converting the region integrals into boundary integrals. We also define a larger class of region functionals based on the estimation and comparison to a prototype of the probability density distribution of image features and show how the shape derivative tool allows us to easily compute the corresponding Gâteaux derivatives and Euler--Lagrange equations. Finally we apply this new functional to the problem of regions segmentation in sequences of color images. We briefly describe our numerical scheme and show some experimental results.

Abstract—Independent component analysis (ICA) aims at decomposing an observed random vector into statistically independent variables. Deflation-based implementations, such as the popular one-unit FastICA algorithm and its variants, extract the independent components one after another. A novel method for deflationary ICA, referred to as RobustICA, is put forward in this paper. This simple technique consists of performing exact line search optimization of the kurtosis contrast function. The step size leading to the global maximum of the contrast along the search direction is found among the roots of a fourth-degree polynomial. This polynomial rooting can be performed algebraically, and thus at low cost, at each iteration. Among other practical benefits, RobustICA can avoid prewhitening and deals with real- and complex-valued mixtures of possibly noncircular sources alike. The absence of prewhitening improves asymptotic performance.

With heart and cardiovascular diseases continually challenging healthcare systems worldwide, translating basic research on cardiac (patho)physiology into clinical care is essential. Exacerbating this already extensive challenge is the complexity of the heart, relying on its hierarchical structure and function to maintain cardiovascular flow. Computational modelling has been proposed and actively pursued as a tool for accelerating research and translation. Allowing exploration of the relationships between physics, multiscale mechanisms and function, computational modelling provides a platform for improving our understanding of the heart. Further integration of experimental and clinical data through data assimilation and parameter estimation techniques is bringing computational models closer to use in routine clinical practice. This article reviews developments in computational cardiac modelling and how their integration with medical imaging data is providing new pathways for translational cardiac modelling.

This paper deals with the image-based visual servoing (IBVS), subject to constraints. Robot workspace limitations, visibility constraints, and actuators limitations are addressed. These constraints are formulated into state, output, and input constraints, respectively. Based on the predictive-control strategy, the IBVS task is written into a nonlinear optimization problem in the image plane, where the constraints can be easily and explicitly taken into account. Second, the contribution of the image prediction and influence of the prediction horizon are pointed out. The image prediction is obtained due to a model. The latter can be a local model based on the interaction matrix or a nonlinear global model based on 3-D data. Its choice is discussed with respect to the constraints to be handled. Finally, simulations that were obtained with a 6-degree-of-freedom (DOF) free-flying camera highlight the potential advantages of the proposed approach with respect to the image prediction and the constraint handling.

Several modifications have been proposed to speed up the alternating least squares (ALS) method of fitting the PARAFAC model. The most widely used is line search, which extrapolates from linear trends in the parameter changes over prior iterations to estimate the parameter values that would be obtained after many additional ALS iterations. We propose some extensions of this approach that incorporate a more sophisticated extrapolation, using information on nonlinear trends in the parameters and changing all the parameter sets simultaneously. The new method, called “enhanced line search (ELS),” can be implemented at different levels of complexity, depending on how many different extrapolation parameters (for different modes) are jointly optimized during each iteration. We report some tests of the simplest parameter version, using simulated data. The performance of this lowest-level of ELS depends on the nature of the convergence difficulty. It significantly outperforms standard LS when there is a “convergence bottleneck,” a situation where some modes have almost collinear factors but others do not, but is somewhat less effective in classic “swamp” situations where factors are highly collinear in all modes. This is illustrated by examples. To demonstrate how ELS can be adapted to different N-way decompositions, we also apply it to a four-way array to perform a blind identification of an under-determined mixture (UDM). Since analysis of this dataset happens to involve a serious convergence “bottleneck” (collinear factors in two of the four modes), it provides another example of a situation in which ELS dramatically outperforms standard line search.