Gesture recognition is widely used to express emotions or to communicate with other people or machines. Hand gesture recognition is a problem of great interest to researchers because it is a high-dimensional pattern recognition problem. The high dimensionality of the problem is directly related to the performance of machine learning models. The dimensionality problem can be addressed through feature selection and feature extraction. In this sense, the evaluation of a model with manual feature extraction and automatic feature extraction was proposed. The manual feature extraction was performed using the statistical functions of central tendency, while the automatic extraction was performed by means of a CNN and BiLSTM. These features were also evaluated in classifiers such as Softmax, ANN, and SVM. The best-performing model was the combination of BiLSTM and ANN (BiLSTM-ANN), with an accuracy of 99.9912%.

Hand gestures are widely used in human-to-human and human-to-machine communication. Therefore, hand gesture recognition is a topic of great interest. Hand gesture recognition is closely related to pattern recognition, where overfitting can occur when there are many predictors relative to the size of the training set. Therefore, it is necessary to reduce the dimensionality of the feature vectors through feature selection techniques. In addition, the need for portability in hand gesture recognition systems limits the use of deep learning algorithms. In this sense, a study of feature selection and extraction methods is proposed for the use of traditional machine learning algorithms. The feature selection methods analyzed are: maximum relevance and minimum redundancy (MRMR), Sequential, neighbor component analysis without parameters (NCAsp), neighbor component analysis with parameters (NCAp), Relief-F, and decision tree (DT). We also analyze the behavior of feature selection methods using classification and recognition accuracy and processing time. Feature selection methods were fed through seventeen feature extraction functions, which return a score proportional to its importance. The functions are then ranked according to their scores and fed to machine learning algorithms such as Artificial Neural Networks (ANN), Support Vector Machine (SVM), K-Nearest Neighbor (KNN), and Decision Tree (DT). This work demonstrates that all feature selection methods evaluated on ANN provide better accuracy. In addition, the combination and number of feature extraction functions influence the accuracy and processing time.

People with physical disabilities (PD) have problems carrying out daily activities, affecting their independence. In this context, a person with PD can communicate through hand gestures or facial gestures, among others. However, selecting the features and patterns that separate one gesture from another is not a trivial problem. In this sense, we propose a real-time domotic system (DS) that works with three subsystems. The first subsystem recognizes hand gestures using a machine learning model and infrared information. The machine learning model consists of pre-processing, feature extraction, classification and post-processing modules. The second subsystem relays the message between the subsystems. Finally, the third subsystem activates the operation of the actuators by gestures. The Hand Gesture Recognition (HGR) model was trained using 6720 observations and tested offline with 1680 observations, giving an accuracy rate of 92.759

This paper presents a proposal for real-time hand gesture recognition for both dynamic and static gestures. For gesture detection, a dataset is obtained using capture sensors such as Leap Motion and Myo armband. The training process is managed by stages of data acquisition, preprocessing, feature extraction, and classification. In this context, in the classification stage, we determine the machine learning algorithms studies through mathematical model investigations. These algorithms allow to determine the real-time processing of hand gesture types and transfer them to a KUKA youBot for control. An application programming interface with ROS (Robot Operating System) is used for the communication process between the sensors and the robot. Finally, a hand gesture control environment is created after obtaining the training data set for the control of the KUKA youBot to test the remote operation with machine learning algorithms.

Hand gestures are widely used in communication among people or between people and machines. Therefore, hand gesture recognition is a topic of great interest. Hand gesture recognition is closely associated with pattern recognition, where overfitting may occur when dealing with many predictors relative to the size of the training set. Therefore, it is necessary to reduce the dimensionality of the feature vectors through feature selection techniques. This study compares three feature selection methods: maximum relevance and minimum redundancy (mrmr), neighbor’s component analysis (nca), and Relief-F, the nca method is analyzed in two variants. We analyze the behavior of these feature selection methods using the classification and recognition accuracy and time of processing. Feature selection methods were fed through seventeen feature-extraction functions, returning a score proportional to its importance. Next, the functions are ordered according to the score and fed to the machine learning algorithms, such as ANN, SVM, KNN, and DT. This work demonstrates that the classification and recognition accuracy have significant growth using the first three or four feature-extraction functions, then stabilized and finally decreased. Also, this work demonstrate that the processing time presents a linear growth according to the number of feature extraction functions used.

This paper presents a systematic review of the literature for the selection of a mathematical model to determine the behavior and constitution of electromyography (EMG) signals. The selection of a mathematical model can be determined according to the search criteria related to obtaining the same EMG signal. In this context, determining the alternatives and most representative characteristics of a mathematical model is done through a hierarchical analytical process (AHP). This process allows determining the functional, modular, descriptive, and resulting characteristics of a proposed mathematical model. Subsequently, these criteria are correlated with quantitative values for each alternative by means of the Centroid Method, which allows for multiple criteria decision making. Finally, the study presenting the mathematical models proposed by the EMG composition with the highest acceptance weight for Hand Gesture Recognition using Electromyographic Signals has been performed.

Hand gesture recognition (HGR) systems are the current topic, attracting interest in many fields. This broad interest is because people use hand movements to communicate and interact with the physical world. HGR systems are overgrowing, and the reason is that they have applications for different fields of study. Fields can be human-computer interaction (HIC), augmented and virtual reality, robotics, medicine, and video games. Recognizing the frames to correspond to the hand gesture from a frames sequence is essential to developing HIC systems. Thus, this paper presents algorithms to detect the images corresponding to a hand gesture from a frame sequence acquired by the Leap Motion Controller. The frames sequence contains non-gestures images because the movement follows a video pattern in which the initial and final images correspond to the transition of the gesture. Therefore, this paper develops an automatic (AID) and manual (MID) images discriminator. Every algorithm returns a dataset with images corresponding to the hand gesture. To validate the algorithms, we present an HGR model with every algorithm. The models take as input the new dataset and feed an architecture based on convolutional neural networks (CNN). Our models recognize five static gestures: open hand, fist, wave in, wave out and pinch. The results show a classification accuracy of 92.31% with MID and 94.70% with AID.

This article presents a system for improving people’s cognitive abilities using electromyography (EMG) signals. This system was created under the serious game mode with a new way of interaction with the user. The sensor Myo Armband was used to measure the EMG signals, which sends the signals via Bluetooth to a Hand Gesture Recognition Model based on a Support Vector Machine to detect the gesture made by the player. This Hand-Gesture Recognition Model was developed in Matlab 2020a. The correct commands given by the recognition model are sent to another computer through the TCP / IP communication protocol to visualize the effects on the game character. The game was developed with the Unity 3D graphics engine to generate an application that optimizes resources. The objective of the game is to solve mazes with two levels of difficulty which allow improving cognitive skills such as attention, concentration, observation, abstract thinking, psycho-motor skills and problem solving.

The research consists of hand gesture recognition using a commercial sensor called Leap Motion Controller. For the study, we determined hand tracking carried out for 5 static and 4 dynamic gestures. For this work, the stages of the gesture recognition study were divided into data acquisition, preprocessing, feature extraction, and classification. For the data acquisition stage, we programmed an interface in MATLAB, which generates a structure of output data in images and spatial position forms. The preprocessing consists of taking the data acquired by the leap motion controller and transform them into data readable by MATLAB using the MEX compiler, these data are captured during the execution of the gesture. Besides, the proposed data acquisition interface collect demographic data, hand injuries, and the light intensity of the environment in which the data is taken. These data can be used in hand gesture recognition systems and application fields such as medicine, engineering, and robotics, among others. The development of the acquisition system is parameterized with 30 repetitions at 60 frames per second and the sampling time at 5 s. Finally, the system previews a graph of the gesture recognition test structure in specific activity time.

This article presents a systematic study for the selection of classifiers for hand gesture recognition by electromyography signals. The selection of a classifier can be determined using an arbitrary search criterion or employing an Analytic Hierarchy Process (AHP). The classifiers are determined as alternatives, which for the study are K-Nearest Neighbors (KNN), Naïve Bayes (NB), Support Vector Machines (SVM), and Linear Discriminant Analysis (LDA). Each of these alternatives is selected through search criteria, which are average training accuracy, average testing accuracy, sampling fee, average acceptance rating, and scrap rating average. These criteria and alternatives allow through a systematic study and the AHP model to select each search factor of the analysis of the systematic mapping (SMS) quantitatively. The results determine the KNN and SVM classifiers as the most used for research projects in electromyography signal recognition (EMG).

The advances in artificial intelligence have been boosting the development of applications in different fields. Some of these fields work in real-time and based on development and implementation in machine learning models. The problems presented by these fields of study involve pattern recognition, which consists of selecting a label and specify the instant of time that defines the gesture. In this context, the paper presents a specific model of real-time hand gesture recognition using the leap motion controller and machine learning algorithms. This model recognizes five static gestures. The gestures are open hand, fist, wave in, wave out, and pinch. The modules used to develop the model are data acquisition, pre-processing, feature extraction, classification, and postprocessing. The input signals to the model are the spatial positions and the direction. These signals are pre-processed by a translation and rotation matrix. Also, these are normalized and smoothed by a Butterworth filter. For the feature extraction module, we use the windows division technique. The model was tested with a window of 7 and 10 features. This signal is delivered to the KNN classifier. It tested the model with k = 1, and k = 3, and with DTW as a distance metric. We are varying the signal warping parameter in w = 1, 5, and w = 7. Finally, the model report 92.22

The research consists in the study of the problem arises from the improper use of software, which triggers an excessive investment in tools that do not represent an academic benefit for an institution of higher education. This work proposes the development of an adaptive system which captures the data generated from the execution of any computer application in the laboratories. At the same time, the modules send the data to a server hosted in the cloud where the activities are stored in offline files. On the other hand, the monitoring system tabulates these data and presents them through reports following the XP methodology to facilitate their maintenance. The implementation of the monitoring system made it possible to determine the use of the applications executed in a given time, for the proper control of academic systems.

The research consists in the hand gestures are movements that convey information and thus complement oral communication or by themselves constitute a form of communication between people. The function of a hand gesture recognition system is to identify the type of movement, from a given set of movements, and the instant when that movement is performed. Gesture recognition systems have multiple applications including sign language translation, bionics, human-machine interaction, gamming, and virtual reality. For this reason, hand gesture recognition is a problem where many researchers have focused their attention too. In this context, in this paper, we present a systematic literature review for hand gesture recognition using ma-chine learning and infrared information. This work has been made because there is no work in the scientific literature that reviews gesture recognition systems based on machine learning and infrared information. In this work, we answer the research question: what is the architecture of the proposed models for hand gesture recognition based on machine learning and infrared information? For answering this research question, we used Kitchenham methodology. Finally, in this work, we also present trends and gaps with respect to the problem analyzed.

Currently, gesture recognition is presented as a problem of feature extraction and pattern recognition, in which we label a movement as belonging to a given class. The response of a gesture recognition system can be applied for different problems in different fields, such as medicine, teaching, robotics among others. There are different models proposed in the scientific literature that try to solve the problem of hand gesture recognition. These works do not meet the demand for real-time processing and high recognition accuracy, simultaneously. In this context, the present work describes a project to develop a new model for the recognition of hand gestures using infrared information, acquired with the Leap Motion Controller, and also using machine learning techniques. The proposed model is intended to recognize 5 static and 4 dynamic gestures of the hand in real-time and with high accuracy, simultaneously. The methodology that will be used for developing this work is composed of two phases, the training and testing: In the first phase, we will design, classify, preliminary validation and tune the proposed model. In the second phase, we will test the proposed model by estimating its accuracy of recognition and time of processing.

El Estado ecuatoriano está interesado en promover la integración de energías renovables en el sector transporte, con especial atención a las islas Galápagos. En este contexto, el presente artículo describe el diseño e implementación de una plataforma informática cloud de monitoreo del comportamiento de una estación de recarga de vehículos eléctricos, alimentada desde un sistema solar fotovoltaico aislado, actualmente en funcionamiento en el campus Huachi de la Universidad Técnica de Ambato. En el presente estudio se realiza una breve descripción del diseño, dimensionado, instalación y puesta en marcha de la electrolinera solar fotovoltaica instalada en la ciudad de Ambato. El monitoreo remoto de la electrolinera está integrado en una plataforma cloud que garantiza realizar el seguimiento del óptimo funcionamiento de instalaciones solares en diferentes zonas geográficas del país. En este sentido, la electrolinera solar fotovoltaica representa un prototipo replicable y escalable en zonas aisladas, en especial en las islas Galápagos. La plataforma cloud de monitoreo remoto permite analizar el comportamiento de los diferentes elementos de la instalación solar: paneles, reguladores, baterías e inversores. La presentación de los datos adquiridos de forma visual permitirá al personal a cargo de la instalación registrar el correcto comportamiento de la electrolinera y estar alerta ante fallos inesperados, garantizando el normal funcionamiento de la electrolinera.

In this article, the design and implementation of an innovative photovoltaic system prototype intelligent control lighting based on Arduino and ZigBee described. The lighting system is connected to a power generation system composed of isolated photovoltaic cells. The lighting control system is subject to a vehicle detection permitting variation of the light intensity when vehicles circulating on the highway. The variation of the light intensity and control of the lighting system is managed through an Arduino microcontroller which in turn transmits information wirelessly to other luminaires distributed along the highway. Thus, it is expected that the lighting system present a significant savings in energy consumption by making use of innovative technologies.

This article describes a platform of acquisition, processing and visualization of the behavior for isolated solar systems called SOLAR MANAGER. The recording of measurements was conducted in an isolated photovoltaic installation of the Faculty of Engineering Electronics Systems and Industrial, Ambato Technical University. The acquisition system will provide the information necessary for performing maintenance monitoring, operation and supervision through variables such as voltage, current, temperature and humidity system to generate renewable energy remotely from a CLOUD COMPUTING platform. It is intended that the implemented system is economically viable, because the system is developed in open source allowing compete with commercial monitoring systems. The control and monitoring of photovoltaic installations isolated platform - SOLAR MANAGER, in the future one will become a central monitoring and control of all isolated PV plants installed all over the country