Search

Objective We used deep convolutional neural networks (DCNNs) to classify electroencephalography (EEG) signals in a steady-state visually evoked potentials (SSVEP) based single-channel brain–computer interface (BCI), which does not require calibration on the user. Methods EEG signals were converted to spectrograms and served as input to train DCNNs using the transfer learning technique. We also modified and applied a data augmentation method, SpecAugment, generally employed for speech recognition. Furthermore, for comparison purposes, we classified the SSVEP dataset using support-vector machines (SVMs) and filter bank canonical correlation analysis (FBCCA). Results Excluding the evaluated user's data from the fine-tuning process, we reached 82.2% mean test accuracy and 0.825 mean F1-Score on 35 subjects from an open dataset, using a small data length (0.5 s), only one electrode (Oz) and the DCNN with transfer learning, window slicing (WS) and SpecAugment's time masks. Conclusion The DCNN results surpassed SVM and FBCCA performances, using a single electrode and a small data length. Transfer learning provided minimal accuracy change, but made training faster. SpecAugment created a small performance improvement and was successfully combined with WS, yielding higher accuracies. Significance We present a new methodology to solve the problem of SSVEP classification using DCNNs. We also modified a speech recognition data augmentation technique and applied it to the context of BCIs. The presented methodology surpassed performances obtained with FBCCA and SVMs (more traditional SSVEP classification methods) in BCIs with small data lengths and one electrode. This type of BCI can be used to develop small and fast systems.

Independent Component Analysis over finite fields is an unsupervised signal processing problem that poses a challenging combinatorial optimization task. In this context, a solution based on Simulated Annealing, with an entropy-based objective function, is proposed. The empirical results demonstrate the effectiveness of the method, with a performance competitive or superior to that of the reference techniques, at an inferior asymptotic computational cost.

Agradecimentos: The authors would like to thank São Paulo’s Research Foundation (FAPESP; grants #2016/22116-9, #2017/10341-0 and #2013/07559-3) and the Brazilian’s Council for Research and Development (CNPq; grants #308811/2019-4 and #142229/2016-4) Abstract: Motor imagery (MI) has been associated to clinical benefits for motor and cognitive rehabilitation protocols through the appropriate identification and training of its corresponding neural patterns. Although potentially impactful factors for MI practice have been identified, issues regarding system feedback and the underlying neurophysiological alterations such practice can produce still remain. In this work, we investigated how feedback can affect the outcome of a MI practice protocol with a brain-computer interface (BCI), when providing: no feedback, sham feedback or actual neurofeedback. We assessed whether event-related desynchronization occurrence (EO) and classification accuracy (CA) varied in each scenario. We found that: (i) practice without feedback did not enhance either metric; (ii) sham feedback yielded a wider variety of outcomes than for the control group, but no statistical significance was observed at the group level; (iii) although users trained to adapt to the BCI classifier, improvements in CA and both contra and ipsilateral EO were observed for the neurofeedback group: average CA increased from (71.3 +/- 3.7) % to (75.5 +/- 5.8) %, whereas median EO increased approximately 5% per session This study reinforces the relevance of neurofeedback in MI learning and provides new insights into neurophysiological correlates throughout the practice protocol. We also hint at a possible placebo effect that provides significant correlations between EO and CA only for the sham group FUNDAÇÃO DE AMPARO À PESQUISA DO ESTADO DE SÃO PAULO - FAPESP CONSELHO NACIONAL DE DESENVOLVIMENTO CIENTÍFICO E TECNOLÓGICO - CNPQ Fechado

Agradecimentos: This work was supported by the São Paulo Research Foundation (FAPESP, Brazil – Grant 2013/07559-3), the Studies and Projects Funding Agency(FINEP, Brazil), the National Council of Scientific and Technological Development (CNPq, Brazil), and the Coordination for the Improvement of Higher Education Personnel (CAPES, Brazil) Abstract: The study of motor imagery (MI) has been a subject of great interest within the brain-computer interface (BCI) community. Several approaches have been proposed to solve the problem of classifying cerebral responses due to MI, mostly based on the power spectral density of the mu and beta bands; however, no optimum manner of proceeding through the fundamental steps of a MI-BCI has yet been established. In this work, we explored a relatively novel approach regarding feature generation for a MI-BCI by assuming that functional connectivity patterns of the brain are altered during hand MI. We modelled interactions among EEG electrodes by a graph, extracted metrics from it during left and right hand MI from eight subjects and classified the signals using commonly employed techniques in the BCI community (LDA and SVM). We also compared this approach to the more established method of using the signal power spectral density as the classifier features. With the graph method, we confirmed that only specific electrodes provide relevant information for data classification. A first approach provided maximum average classification rates across all subjects for the graph method of 86% for the mu band and 87% for the beta band. For the PSD method, average rates were of 98% and 99% for the mu and beta bands, respectively. However, a much larger number of features was needed: (130 44) and (273 89) for the mu and beta bands, respectively. Aiming to reproduce these rates using the graph method, pairwise inputs combinations of graph metrics were tested. They proved to be sufficient to obtain essentially the same classification accuracy rates, but with a considerably smaller number of features - about 60 features, for both bands. We thus conclude that the graph method is a feasible option for classification of hand MI signals FUNDAÇÃO DE AMPARO À PESQUISA DO ESTADO DE SÃO PAULO - FAPESP FINANCIADORA DE ESTUDOS E PROJETOS - FINEP CONSELHO NACIONAL DE DESENVOLVIMENTO CIENTÍFICO E TECNOLÓGICO - CNPQ COORDENAÇÃO DE APERFEIÇOAMENTO DE PESSOAL DE NÍVEL SUPERIOR - CAPES Fechado

This paper presents a systematic analysis of a game controlled by a Brain-Computer Interface (BCI) based on Steady-State Visually Evoked Potentials (SSVEP). The objective is to understand BCI systems from the Human-Computer Interface (HCI) point of view, by observing how the users interact with the game and evaluating how the interface elements influence the system performance. The interactions of 30 volunteers with our computer game, named “Get Coins,” through a BCI based on SSVEP, have generated a database of brain signals and the corresponding responses to a questionnaire about various perceptual parameters, such as visual stimulation, acoustic feedback, background music, visual contrast, and visual fatigue. Each one of the volunteers played one match using the keyboard and four matches using the BCI, for comparison. In all matches using the BCI, the volunteers achieved the goals of the game. Eight of them achieved a perfect score in at least one of the four matches, showing the feasibility of the direct communication between the brain and the computer. Despite this successful experiment, adaptations and improvements should be implemented to make this innovative technology accessible to the end user.

In blind channel equalization, the use of criteria from the field of information theoretic learning (ITL) has already proved to be a promising alternative, since the use of the high-order statistics is mandatory in this task. In view of the several existent ITL propositions, we present in this work a detailed comparison of the main ITL criteria employed for blind channel equalization and also introduce a new ITL criterion based on the notion of distribution matching. The analyses of the ITL framework are held by means of comparison with elements of the classical filtering theory and among the studied ITL criteria themselves, allowing a new understanding of the existing ITL framework. The verified connections provide the basis for a comparative performance analysis in four practical scenarios, which encompasses discrete/continuous sources with statistical independence/dependence, and real/complex-valued modulations, including the presence of Gaussian and non-Gaussian noise. The results indicate the most suitable ITL criteria for a number of scenarios, some of which are favorable to our proposition.

Agradecimentos: The authors would like to thank FAPESP (Processes n. 2014/27108-9 and 2015/16325-1) and CNPq (Processes n. 311786/2014-6 and 305621/2015-7) for the financial support Abstract: Several problems in signal processing are addressed by expert systems which take into account a set of priors on the sought signals and systems. For instance, blind source separation is often tackled by means of a mono-objective formulation which relies on a separation criterion associated with a given property of the sought signals (sources). However, in many practical situations, there are more than one property to be exploited and, as a consequence, a set of separation criteria may be used to recover the original signals. In this context, this paper addresses the separation problem by means of an approach based on multi-objective optimization. Differently from the existing methods, which provide only one estimate for the original signals, our proposal leads to a set of solutions that can be utilized by the system user to take his/her decision. Results obtained through numerical experiments over a set of biomedical signals highlight the viability of the proposed approach, which provides estimations closer to the mean squared error solutions compared to the ones achieved via a mono-objective formulation. Moreover, since our proposal is quite general, this work also contributes to encourage future researches to develop expert systems that exploit the multi-objective formulation in different source separation problems CONSELHO NACIONAL DE DESENVOLVIMENTO CIENTÍFICO E TECNOLÓGICO - CNPQ FUNDAÇÃO DE AMPARO À PESQUISA DO ESTADO DE SÃO PAULO - FAPESP Fechado

In this work, a Brain-Computer Interface based on motor imagery was implemented focusing on the signal processing stage. Methods based on machine learning were used along with feature selection techniques on the treatment and classification of data.

International audience; In the context of nonlinear Blind Source Separation (BSS), the Post-Nonlinear (PNL) model is of great importance due to its suitability for practical nonlinear problems. Under certain mild constraints on the model, Independent Component Analysis (ICA) methods are valid for performing source separation, but requires use of Higher-Order Statistics (HOS). Conversely, regarding the sole use of the Second-Order Statistics (SOS), their study is still in an initial stage. In that sense, in this work, the conditions and the constraints on the PNL model for SOS-based separation are investigated. The study encompasses a time-extended formulation of the PNL problem with the objective of extracting the temporal structure of the data in a more extensive manner, considering SOS-based methods for separation, including the proposition of a new one. Based on this, it is shown that, under some constraints on the nonlinearities and if a ✩ This work was partially supported by CNPq (305621/2015-7), FAPESP given number of time delays is considered, source separation can be successfully achieved, at least for polynomial nonlinearities. With the aid of metaheuristics called Differential Evolution and Clonal Selection Algorithm for optimization, the performances of the SOS-based methods are compared in a set of simulation scenarios, in which the proposed method shows to be a promising approach.

A Brain Computer Interface (BCI) provides a channel of communication and control amid the human brain and a computer or electronic device. In this paper we describe the implementation of all the main steps involved in the operation of a BCI system based on Steady State Visually Evoked Potential (SSVEP) paradigm. Also, we present a comparative study about the performance of our BCI employing two feature extraction techniques (Welch’s Method and Discrete Fourier Transform), and considering the signal analysis window of two sizes (4 and 6 s). The results guided the conception of an online BCI-SSVEP. Our system was successfully controlled by two volunteers, who completed the task of moving a foam block with a robotic manipulator.

The problem of blind source separation (BSS) has been intensively studied by the signal processing community. The first solutions to deal with BSS were proposed in the 1980's and are founded on the concept of independent component analysis (ICA). More recently, aiming at tackling some limitations of ICA-based methods, much attention has been paid to alternative BSS approaches. In this tutorial, in addition to providing a brief review of the classical BSS framework, we present two research trends in this area, namely source separation over Galois fields and sparse component analysis. For both subjects, we provide an overview of the main criteria, highlighting scenarios that can benefit from these more recent BSS paradigms.

This work proposes a new approach to the blind inversion of Wiener systems. A Wiener system is composed of a linear time-invariant (LTI) sub-system followed by a memoryless nonlinear function. The goal is to recover the input signal by knowing just the output of the Wiener system, and the straightforward scheme to achieve this is called the Hammerstein system - apply a memoryless nonlinear mapping followed by a LTI sub-system to the output signal of the Wiener system. If the input of the Wiener system is originally iid and some mild conditions are satisfied, the inversion is possible. Based on this statement and the limitations of relevant previous works, a solution is proposed combining (i) immune-inspired optimization algorithms, (ii) information theory and (iii) IIR filters that yield a robust scheme with a relatively reduced risk of local convergence. Experimental results indicated a similar or superior performance of the new approach, in comparison with two other blind methodologies. HighlightsThis work proposes an immune-inspired algorithm to perform blind inversion of Wiener systems.The method employs mutual information-based criteria and IIR filters to adapt the inverse structure.The experimental results indicate a superior or equivalent performance of the new technique, in comparison with gradient-based search or kurtosis-based criterion.

Machine learning has been increasingly used in current days. Great improvements, especially in deep neural networks, helped to boost the achievable performance in computer vision and signal processing applications. Although different techniques were applied for deep architectures, the frequency domain has not been thoroughly explored in this field. In this context, this paper presents a new method for extracting discriminative features according to the Fourier analysis. The proposed frequency extractor layer can be combined with deep architectures in order to improve image classification. Computational experiments were performed on face liveness detection problem, yielding better results than those presented in the literature for the grandtest protocol of Replay-Attack Database. This paper also aims to raise the discussion on how frequency domain layers can be used in deep architectures to further improve the network performance.

Signal processing methods based on blind source separation (BSS) have been successfully applied to deal with the interference problem in ion-selective electrode (ISE) arrays. So far, the development of BSS methods to ISE arrays strongly relied on the Nicolsky-Eisenman (NE) model. Since this model provides a rough approximation in certain situations, the present paper aims at improving the existing BSS solutions by refining the mixing model based on the NE equation. More precisely, we place a set of polynomial monotonic functions after the standard logarithmic functions within the NE. The obtained results are preliminary but attest that the proposal lead to slight improvements and open the way for investigations on alternative monotonic compensating functions.

In this work, we present a novel bioinspired framework for performing ICA over finite (Galois) fields of prime order P. The proposal is based on a state-of-the-art immune-inspired algorithm, the cob-aiNet[C], which is employed to solve a combinatorial optimization problem - associated with a minimal entropy configuration - adopting a Michigan-like population structure. The simulation results reveal that the strategy is capable of reaching a performance similar to that of standard methods for lower-dimensional instances with the advantage of also handling scenarios with an elevated number of sources.

In 2007, a theory of ICA over finite fields emerged and an algorithm based on pairwise comparison of mixtures, called MEXICO, was developed to deal with this new problem. In this letter, we propose improvements in the method that, according to simulations in GF(2) and GF(3) scenarios, lead to a faster convergence and better separation results, increasing the application possibilities of the new theory in the context of large databases.

This paper addresses the problem of direction-of-arrival (DOA) parameter estimation in array processing when the signals are inherently discrete, which is the case mainly in the digital communication context. Based on the particular structure of the signal space in the data model, a maximum likelihood-based approach is introduced. The strategy consists in transforming the parameter estimation problem into a decision task. It is shown through numerical simulations that the proposed solution closely follows the performance limit given by the Cramer–Rao bound. Some important features of the technique are as follows: (i) it is capable of handling any number of sources, provided that the number of sensors is greater than or equal to two and the number of snapshots is sufficiently greater than the cardinality of the signal space; (ii) the estimation quality is not affected by the angle and phase separation; and (iii) it offers the possibility to deal with uncalibrated arrays.

We propose a novel strategy to generate initial candidate solutions for bio-inspired algorithms applied to the direction of arrival estimation problem. The idea, which aims to improve the efficiency of the estimator, consists in using the frequency response of a well-known optimum noise reduction filter as the probability density function of the set of candidate solutions. In accordance to this approach, we also employ a modified likelihood function to reduce the estimation error. Simulation results considering an immune-inspired algorithm confirm a significant improvement of its performance and efficiency, and the new estimator reaches the conditional Cramer–Rao lower bound.

International audience; In this paper, we address the problem of blind compensation of nonlinear distortions. Our approach relies on the assumption that the input signals are bandlimited. We then make use of the classical result that the output of a nonlinear memoryless system has a wider spectrum than the one of the in- put signal. However, different from previous works, our approach does not assume the knowledge of the input signal's bandwidth. The proposed approached is considered in the development of a two-stage method for blind source separation (BSS) in post- nonlinear (PNL) models. Indeed, once the functions present in the nonlinear stage of a PNL model are compensated, one can apply the well-established linear BSS algorithms to complete the task of separating the sources. Numerical experiments performed in different scenarios attest the viability of the proposal. Moreover, the proposed method is tested in a real situation where the data are acquired by smart chemical sensor arrays.

This work presents a study of the performance of populational meta-heuristics belonging to the field of natural computing when applied to the problem of direction of arrival (DOA) estimation, as well as an overview of the literature about the use of such techniques in this problem. These heuristics offer a promising alternative to the conventional approaches in DOA estimation, as they search for the global optima of the maximum likelihood (ML) function in a framework characterized by an elegant balance between global exploration and local improvement, which are interesting features in the context of multimodal optimization, to which the ML-DOA estimation problem belongs. Thus, we shall analyze whether these algorithms are capable of implementing the ML estimator, i.e., finding the global optima of the ML function. In this work, we selected three representative natural computing algorithms to perform DOA estimation: differential evolution, clonal selection algorithm, and the particle swarm. Simulation results involving different scenarios confirm that these methods can reach the performance of the ML estimator, regardless of the number of sources and/or their nature. Moreover, the number of points evaluated by such methods is quite inferior to that associated with a grid search, which gives support to their application.

This work aims to present a new method to perform blind extraction of chaotic deterministic sources mixed with stochastic signals. This technique employs the recurrence quantification analysis (RQA), a tool commonly used to study dynamical systems, to obtain the separating system that recovers the deterministic source. The method is applied to invertible and underdetermined mixture models considering different stochastic sources and different RQA measures. A brief discussion about the influence of recurrence plot parameters on the robustness of the proposal is also provided and illustrated by a set of representative simulations.

In this paper we propose a blind source separation method to process the data acquired by an array of ion-selective electrodes in order to measure the ionic activity of different ions in an aqueous solution. While this problem has already been studied in the past, the method presented differs from the ones previously analyzed by approximating the mixing function by a second-degree polynomial, and using a method based on the differential of the mutual information to adjust the parameter values. Experimental results, both with synthetic and real data, suggest that the algorithm proposed is more accurate than the other models in the literature.

This work proposes a framework to determine the optimal Wiener equalizer by using an artificial immune network model together with the constant modulus (CM) cost function. This study was primarily motivated by recent theoretical results concerning the CM criterion and its relation to the Wiener approach. The proposed immune-based technique was tested under different channel models and filter orders, and benchmarked against a procedure using a genetic algorithm with niching. The results demonstrated that the proposed strategy has a clear superiority when compared with the more traditional technique. The proposed algorithm presents interesting features from the perspective of multimodal search, being capable of determining the optimal Wiener equalizer in most runs for all tested channels.

Brain–computer interface (BCI) systems based on electroencephalography have been increasingly used in different contexts, engendering applications from entertainment to rehabilitation in a non-invasive framework. In this study, we perform a comparative analysis of different signal processing techniques for each BCI system stage concerning steady state visually evoked potentials (SSVEP), which includes: (1) feature extraction performed by different spectral methods (bank of filters, Welch's method and the magnitude of the short-time Fourier transform); (2) feature selection by means of an incremental wrapper, a filter using Pearson's method and a cluster measure based on the Davies–Bouldin index, in addition to a scenario with no selection strategy; (3) classification schemes using linear discriminant analysis (LDA), support vector machines (SVM) and extreme learning machines (ELM). The combination of such methodologies leads to a representative and helpful comparative overview of robustness and efficiency of classical strategies, in addition to the characterization of a relatively new classification approach (defined by ELM) applied to the BCI-SSVEP systems.

An interesting and recent application of population-based metaheuristics resides in an unsupervised signal processing task: independent component analysis (ICA) over finite fields. Based on a state-of-the-art immune-inspired method, this work proposes a new ICA algorithm for finite fields of arbitrary order that employs mutation and local search operators specifically customized to the problem domain. The results obtained with the new technique indicate that the proposal is effective in performing component separation, and the analysis includes a preliminary study on image separation.

This paper analyzes the application of a multi-population Gaussian-based estimation of distribution algorithm equipped with a restarting strategy and mutation, named MGcEDA, to the problem of estimating the Direction of Arrival (DOA) of time-varying plane waves impinging on a uniform linear array of sensors. This problem requires the minimization of a dynamic cost function which is non-linear, non-quadratic, multimodal and variant with respect to the signal-to-noise ratio. Experiments showed that MGcEDA was able to quickly respond to changes in the source features in scenarios with different levels of noise and number of signals. Moreover, MGcEDA outperforms a previously proposed approach in all considered experiments in terms of well known performance measures.

Echo state networks (ESNs) characterize an attractive alternative to conventional recurrent neural network (RNN) approaches as they offer the possibility of preserving, to a certain extent, the processing capability of a recurrent architecture and, at the same time, of simplifying the training process. However, the original ESN architecture cannot fully explore the potential of the RNN, given that only the second-order statistics of the signals are effectively used. In order to overcome this constraint, distinct proposals promote the use of a nonlinear readout aiming to explore higher-order available information though still maintaining a closed-form solution in the least-squares sense. In this work, we review two proposals of nonlinear readouts - a Volterra filter structure and an extreme learning machine - and analyze the performance of these architectures in the context of two relevant signal processing tasks: supervised channel equalization and chaotic time series prediction. The obtained results reveal that the nonlinear readout can be decisive in the process of aproximating the desired signal. Additionally, we discuss the possibility of combining both ideas of nonlinear readouts and preliminary results indicate that a performance improvement can be attained.

Echo state networks (ESNs) can be interpreted as promoting an encouraging compromise between two seemingly conflicting objectives: (i) simplicity of the resulting mathematical model and (ii) capability to express a wide range of nonlinear dynamics. By imposing fixed weights to the recurrent connections, the echo state approach avoids the well-known difficulties faced by recurrent neural network training strategies, but still preserves, to a certain extent, the potential of the underlying structure due to the existence of feedback loops within the dynamical reservoir. Moreover, the overall training process is relatively simple, as it amounts essentially to adapting the readout, which usually corresponds to a linear combiner. However, the linear nature of the output layer may limit the capability of exploring the available information, since higher-order statistics of the signals are not taken into account. In this work, we present a novel architecture for an ESN in which the linear combiner is replaced by a Volterra filter structure. Additionally, the principal component analysis technique is used to reduce the number of effective signals transmitted to the output layer. This idea not only improves the processing capability of the network, but also preserves the simplicity of the training process. The proposed architecture is then analyzed in the context of a set of representative information extraction problems, more specifically supervised and unsupervised channel equalization, and blind separation of convolutive mixtures. The obtained results, when compared to those produced by already proposed ESN versions, highlight the benefits brought by the novel network proposal and characterize it as a promising tool to deal with challenging signal processing tasks.

In this work, we present a discussion concerning some fundamental aspects of sparse component analysis (SCA), a methodology that has been increasingly employed to solve some challenging signal processing problems. In particular, we present some insights into the use of l1 norm as a quantifier of sparseness and its application as a cost function to solve the blind source separation (BSS) problem. We also provide results on experiments in which source extraction was successfully made when we performed a search for sparse components in the mixtures of sparse signals. Finally, we make an analysis of the behavior of this approach on scenarios in which the source signals are not sparse.

We propose a method for source separation of convolutive mixture based on nonlinear prediction-error filters. This approach converts the original problem into an instantaneous mixture problem, which can be solved by any of the several existing methods in the literature. We employ fuzzy filters to implement the prediction-error filter, and the efficacy of the proposed method is illustrated by some examples.

In light of the recently proposed generalized correlation function named correntropy, which exploits higher-order statistics and the time structure of signals, we have, in this work, two main objectives: 1) to give a new interpretation --- founded on the relationships between the constant modulus (CM) and Shalvi-Weinstein criteria and between the latter and methods for ICA based on nongaussianity --- to the performance of the constant modulus approach under dependent sources and 2) to analyze the correntropy in the context of blind deconvolution of i.i.d. and dependent sources, as well as to establish elements of a comparison between it and the CMA. The analyses and simulation results unveil some theoretical aspects hitherto unexplored.

This work analyzes representative algorithms for adaptive equalization from the standpoint of Lyapunov stability. The proposed approach is valid for both supervised and unsupervised algorithms. Using this framework, we were able to obtain results that are applicable to the entire parameter space, and that, for certain conditions, are equivalent to bounds previously derived for the steepest descent and for the one-tap CMA. The proposed methodology also leads to a novel analysis of the maximum step-size for the N-tap CMA.