Many studies applying Brain-Computer Interfaces (BCIs) based on Motor Imagery (MI) tasks for rehabilitation have demonstrated the important role of detecting the Event-Related Desynchronization (ERD) to recognize the user's motor intention. Nowadays, the development of MI-based BCI approaches without or very few calibration stages session-by-session for different days or weeks is still an open and emergent scope. In this work, a new scheme is proposed by applying Convolutional Neural Networks (CNN) for MI classification, using an end-to-end Shallow architecture that contains two convolutional layers for temporal and spatial feature extraction. We hypothesize that a BCI designed for capturing event-related desynchronization/synchronization (ERD/ERS) at the CNN input, with an adequate network design, may enhance the MI classification with less calibration stage. The proposed system using the same architecture was tested on three public datasets through multiple experiments, including both subject-specific and nonsubject-specific training. Comparable and also superior results with respect to the state-of-the-art were obtained. On subjects whose EEG data were never used in the training process, our scheme also achieved promising results with respect to existing non-subject-specific BCIs, which is other progress to facilitate the clinical application.
SummaryEven with significant progresses in the maximum power point (MPP) research area, the necessity to improve the existing methods becomes mandatory to increase the energy conversion efficiency. Since the power–voltage (P‐V) characteristic curve of photovoltaic (PV) arrays has multiple peaks under partially shaded (PS) conditions, the conventional MPP tracking (MPPT) control methods have the difficult challenge of locate the global MPP (GMPP) among many local MPPs (LMPPs). In recent years, numerous research papers have been focused on techniques to efficiently track the GMPP and alleviate the partial shading effects. One of the most popular evolutionary search technique is particle swarm optimization (PSO) that provides high tracking speed and the ability operate under different environmental conditions. For solving some conventional PSO technique common weaknesses, several modifications and improvements have emerged in the past years. This paper provides a comparative and comprehensive review of some relevant PSO‐based methods taking into account the effects of important key issues such as particles initialization criteria, search space, convergence speed, initial parameters, performance with and without partial shading, and efficiency. The simulation results are validated under numerous test conditions using MATLAB code and Simulink package.
X-ray methods have proven to be reliable, accurate and sensitive techniques to study activated carbons. The studying of granular activated carbon (GAC) samples through X-ray digital radiographic images using Deep Learning, more specifically convolutional neural networks (CNN) class of model, has been explored. Results were compared to hand-engineered characterization using X-Ray absorption method (XRA). It was proved that CNNs represent a fast and reliable analytical tool for indirect information on the chemical and physical characteristics of GACs. The proposed method opens possibilities for the application of Deep Learning based models on radiographic images for the characterization and comparison of exhausted and virgin porous materials.
Motor Imagery (MI)-based Brain–Computer Interfaces (BCIs) have been widely used as an alternative communication channel to patients with severe motor disabilities, achieving high classification accuracy through machine learning techniques. Recently, deep learning techniques have spotlighted the state-of-the-art of MI-based BCIs. These techniques still lack strategies to quantify predictive uncertainty and may produce overconfident predictions. In this work, methods to enhance the performance of existing MI-based BCIs are proposed in order to obtain a more reliable system for real application scenarios. First, the Monte Carlo dropout (MCD) method is proposed on MI deep neural models to improve classification and provide uncertainty estimation. This approach was implemented using Shallow Convolutional Neural Network (SCNN-MCD) and with an ensemble model (E-SCNN-MCD). As another contribution, to discriminate MI task predictions of high uncertainty, a threshold approach is introduced and tested for both SCNN-MCD and E-SCNN-MCD approaches. The BCI Competition IV Databases 2a and 2b were used to evaluate the proposed methods for both subject-specific and non-subject-specific strategies, obtaining encouraging results for MI recognition.
ResumenLas técnicas de reducción de ruido son ampliamente utilizadas en la grabación de audio, la edición de imágenes y en el procesamiento de señales industriales. La idea es reconstruir los datos originales a partir de la señal ruidosa suprimiendo toda, o casi toda, la distorsión generada por el ruido inherente a los procesos físicos. En el presente trabajo se realiza una comparación de diferentes métodos de supresión de ruido que se basan en la selección adaptativa del umbral. Estas técnicas han sido usadas extensivamente en el procesamiento de imágenes pero el objetivo de este trabajo es evaluar su rendimiento en la reducción de ruido de señales industriales. En particular se analiza el comportamiento de los métodos Bayes Shrink, Normal Shrink, Modified Shrink y Neight Shrink para la reducción de ruido gaussiano en estas señales. A tales efectos se utilizó un conjunto de señales patrón, que incluye a las señales propuestas por Donoho y otras mediciones representativas obtenidas de procesos reales en las plantas de Níquel cubanas. Las pruebas realizadas revelan que el algoritmo Neigh Shrink es el que mejor se comporta en los datos analizados.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.