Search

Gestures are one of the most important agents for human-computer interaction. They play a mediator role between human intention and the control of machines. Electromyography (EMG) pattern recognition has been studied for gesture recognition for years to control of prostheses and rehabilitation systems. EMG data gives information about the electrical activity related to muscles. It is obtained from the arm and helps to understand hand gestures. For this work, hand gesture data taken from UCI2019 EMG dataset obtained from myo Thalmic armband were classified with six different machine learning algorithms. Artificial Neural Network (ANN), Support Vector Machine (SVM), k-Nearest Neighbor (k-NN), Naive Bayes (NB), Decision Tree (DT) and Random Forest (RF) methods were applied for comparison based on several performance metrics which are accuracy, precision, sensitivity, specificity, classification error, kappa, root mean squared error (RMSE) and correlation. The data belongs to seven hand gestures. 700 samples from 7 classes (100 samples per group) were used in the experiments. Splitting ratio in the classification was 0.8-0.2, i.e. 80% of the samples were used in training and 20% of data were used in the testing phase of the classifier. NB was found to be the best among other methods because of high accuracy. Classification accuracy varies between 97.52% to 100% for each gesture. Considering the results of the performance metrics, it can be said that this study recognizes and classifies seven hand gestures successfully in comparison with the literature. The proposed method can easily be used for human-machine interaction and smart device controlling like prosthesis, wheelchair, and smart entertainment applications. WOS:000704721300002

This work discusses a novel human&ndash, robot interface for a climbing robot for inspecting weld beads in storage tanks in the petrochemical industry. The approach aims to adapt robot autonomy in terms of the operator&rsquo, s experience, where a remote industrial joystick works in conjunction with an electromyographic armband as inputs. This armband is worn on the forearm and can detect gestures from the operator and rotation angles from the arm. Information from the industrial joystick and the armband are used to control the robot via a Fuzzy controller. The controller works with sliding autonomy (using as inputs data from the angular velocity of the industrial controller, electromyography reading, weld bead position in the storage tank, and rotation angles executed by the operator&rsquo, s arm) to generate a system capable of recognition of the operator&rsquo, s skill and correction of mistakes from the operator in operating time. The output from the Fuzzy controller is the level of autonomy to be used by the robot. The levels implemented are Manual (operator controls the angular and linear velocities of the robot), Shared (speeds are shared between the operator and the autonomous system), Supervisory (robot controls the angular velocity to stay in the weld bead, and the operator controls the linear velocity), Autonomous (the operator defines endpoint and the robot controls both linear and angular velocities). These autonomy levels, along with the proposed sliding autonomy, are then analyzed through robot experiments in a simulated environment, showing each of these modes&rsquo, purposes. The proposed approach is evaluated in virtual industrial scenarios through real distinct operators.

In this paper, a customizable wearable 3D-printed bionic arm is designed, fabricated, and optimized for a right arm amputee. An experimental test has been conducted for the user, where control of the artificial bionic hand is accomplished successfully using surface electromyography (sEMG) signals acquired by a multi-channel wearable armband. The 3D-printed bionic arm was designed for the low cost of 295 USD, and was lightweight at 428 g. To facilitate a generic control of the bionic arm, sEMG data were collected for a set of gestures (fist, spread fingers, wave-in, wave-out) from a wide range of participants. The collected data were processed and features related to the gestures were extracted for the purpose of training a classifier. In this study, several classifiers based on neural networks, support vector machine, and decision trees were constructed, trained, and statistically compared. The support vector machine classifier was found to exhibit an 89.93% success rate. Real-time testing of the bionic arm with the optimum classifier is demonstrated.

Surface electromyographic signal (sEMG) is a kind of bioelectrical signal, which records the data of muscle activity intensity. Most sEMG-based hand gesture recognition, which uses machine learning as the classifier, depends on feature extraction of sEMG data. Recently, a deep leaning-based approach such as recurrent neural network (RNN) has provided a choice to automatically learn features from raw data. This paper presents a novel hand gesture prediction method by using an RNN model to learn from raw sEMG data and predict gestures. The sEMG signals of 21 short-term hand gestures of 13 subjects were recorded with a Myo armband, which is a non-intrusive, low cost, commercial portable device. At the start of the gesture, the trained model outputs an instantaneous prediction for the sEMG data. Experimental results showed that the more time steps of data that were known, the higher instantaneous prediction accuracy the proposed model gave. The predicted accuracy reached about 89.6% when the data of 40-time steps (200 ms) were used to predict hand gesture. This means that the gesture could be predicted with a delay of 200 ms after the hand starts to perform the gesture, instead of waiting for the end of the gesture.

Nowadays, gesture-based technology is revolutionizing the world and lifestyles, and the users are comfortable and care about their needs, for example, in communication, information security, the convenience of day-to-day operations and so forth. In this case, hand movement information provides an alternative way for users to interact with people, machines or robots. Therefore, this paper presents a character input system using a virtual keyboard based on the analysis of hand movements. We analyzed the signals of the accelerometer, gyroscope, and electromyography (EMG) for movement activity. We explored potential features of removing noise from input signals through the wavelet denoising technique. The envelope spectrum is used for the analysis of the accelerometer and gyroscope and cepstrum for the EMG signal. Furthermore, the support vector machine (SVM) is used to train and detect the signal to perform character input. In order to validate the proposed model, signal information is obtained from predefined gestures, that is, &ldquo, double-tap&rdquo, &ldquo, hold-fist&rdquo, wave-left&rdquo, wave-right&rdquo, and &ldquo, spread-finger&rdquo, of different respondents for different input actions such as &ldquo, input a character&rdquo, change character&rdquo, delete a character&rdquo, line break&rdquo, space character&rdquo, The experimental results show the superiority of hand gesture recognition and accuracy of character input compared to state-of-the-art systems.

We create an API (Application Programming Interface) for Indonesian Sign Language (Sistem Isyarat Bahasa Indonesia/SIBI) which is called OpenSIBI. In this case study, we use the Myo Armband device to capture hand gesture data movement. It uses five sensors: Accelerometer, Gyroscope, Orientation, Orientation-Euler, and EMG. First, we record, convert and save those data into JSON dataset in the server as data learning. Then, every data request (trial data) from the client will compare them using k-NN Normalization process. OpenSIBI API works as the middleware which integrated to RabbitMQ as the queue request arranger. Every service request from the client will automatically spread to the server with the queue process. As the media observation, we create a client data request by SIBI Words and Alphabeth Game, which allows the user to answer several stages of puzzle-game with Indonesian Sign Language hand gesture. Game-player must use the Myo armband as an interactive device that reads the hand movement and its fingers for answering the questions given. Thus, the data will be classified and normalized by the k-NN algorithm, which will be processed on the server. In this process, data will pass OpenAPI SIBI (which connected to RabbitMQ) to queue every incoming data-request. So, the obtained data will be processed one by one and sent it back to the client as the answer.

Controlling home entertainment devices, like music and video, via an armband could free the user from using remote controls, but assessing their overall usability with mid-air and micro-gestures still represents an open research question today. For this purpose, this paper reports on results gained by jointly conducting and comparing two studies involving participants using a Thalmic Myo armband to control a NetFlix SmartTV and Spotify: (1) a gesture elicitation study to explore a richer set of user-defined gestures, to measure their effectiveness and the user subjective satisfaction of gesture interaction; (2) a System Usability Scale (SUS) to assess the overall usability of this setup and the subjective satisfaction for user-defined gestures.

Hand gesture recognition (HGR) systems using electromyography (EMG) bracelet-type sensors are currently largely used over other HGR technologies. However, bracelets are susceptible to electrode rotation, causing a decrease in HGR performance. In this work, HGR systems with an algorithm for orientation correction are proposed. The proposed orientation correction method is based on the computation of the maximum energy channel using a synchronization gesture. Then, the channels of the EMG are rearranged in a new sequence which starts with the maximum energy channel. This new sequence of channels is used for both training and testing. After the EMG channels are rearranged, this signal passes through the following stages: pre-processing, feature extraction, classification, and post-processing. We implemented user-specific and user-general HGR models based on a common architecture which is robust to rotations of the EMG bracelet. Four experiments were performed, taking into account two different metrics which are the classification and recognition accuracy for both models implemented in this work, where each model was evaluated with and without rotation of the bracelet. The classification accuracy measures how well a model predicted which gesture is contained somewhere in a given EMG, whereas recognition accuracy measures how well a model predicted when it occurred, how long it lasted, and which gesture is contained in a given EMG. The results of the experiments (without and with orientation correction) executed show an increase in performance from 44.5% to 81.2% for classification and from 43.3% to 81.3% for recognition in user-general models, while in user-specific models, the results show an increase in performance from 39.8% to 94.9% for classification and from 38.8% to 94.2% for recognition. The results obtained in this work evidence that the proposed method for orientation correction makes the performance of an HGR robust to rotations of the EMG bracelet.

The surface electromyography signal (sEMG) is widely used for gesture characterization, its reliability is strongly connected to the features extracted from sEMG recordings. This study aimed to investigate the use of two complexity measures, i.e., fuzzy entropy (FEn) and permutation entropy (PEn) for hand gesture characterization. Fourteen upper limb movements, sorted into three sets, were collected on ten subjects and the performances of FEn and PEn for gesture descriptions were analyzed for different computational parameters. FEn and PEn were able to properly cluster the expected numbers of gestures, but computational parameters were crucial for ensuring clusters&rsquo, separability and proper gesture characterization. FEn and PEn were also compared with other eighteen classical time and frequency domain features through the minimum redundancy maximum relevance algorithm and showed the best predictive importance scores in two gesture sets, they also had scores within the subset of the best five features in the remaining one. Further, the classification accuracies of four different feature sets presented remarkable increases when FEn and PEn are included as additional features. Outcomes support the use of FEn and PEn for hand gesture description when computational parameters are properly selected, and they could be useful in supporting the development of robotic arms and prostheses myoelectric control.

The classification of the bio-signal has been used for various purposes in the literature as they are versatile in diagnosis of anomalies, improvement of overall health and sport performance and creating intuitive human computer interfaces. However, automatic identification of the signal patterns on a streaming real-time signal requires a series of complex procedures. A plethora of heuristic methods, such as neural networks and fuzzy systems, have been proposed as a solution. These methods stipulate certain conditions, such as preconditioning the signals, manual feature selection and large number of training samples. In this study, we introduce a novel variant and application of the Collaborative Representation based Classification (CRC) in spectral domain for recognition of hand gestures using raw surface electromyography (EMG) signals. The CRC based methods do not require large number of training samples for an efficient pattern classification. Additionally, we present a training procedure in which a high end subspace clustering method is employed for clustering the representative samples into their corresponding class labels. Thereby, the need for feature extraction and spotting patterns manually on the training samples is obviated. We presented the intuitive use of spectral features via circulant matrices. The proposed Spectral Collaborative Representation based Classification (SCRC) is able to recognize gestures with higher levels of accuracy for a fairly rich gesture set compared to the available methods. The worst recognition result which is the best in the literature is obtained as 97.3% among the four sets of the experiments for each hand gestures. The recognition results are reported with a substantial number of experiments and labeling computation.

The importance of estimating human movement has increased in the field of human motion capture. HTC VIVE is a popular device that provides a convenient way of capturing human motions using several sensors. Recently, the motion of only users&rsquo, hands has been captured, thereby greatly reducing the range of motion captured. This paper proposes a framework to estimate single-arm orientations using soft sensors mainly by combining a Bi-long short-term memory (Bi-LSTM) and two-layer LSTM. Positions of the two hands are measured using an HTC VIVE set, and the orientations of a single arm, including its corresponding upper arm and forearm, are estimated using the proposed framework based on the estimated positions of the two hands. Given that the proposed framework is meant for a single arm, if orientations of two arms are required to be estimated, the estimations are performed twice. To obtain the ground truth of the orientations of single-arm movements, two Myo gesture-control sensory armbands are employed on the single arm: one for the upper arm and the other for the forearm. The proposed framework analyzed the contextual features of consecutive sensory arm movements, which provides an efficient way to improve the accuracy of arm movement estimation. In comparison with the ground truth, the proposed method estimated the arm movements using a dynamic time warping distance, which was the average of 73.90% less than that of a conventional Bayesian framework. The distinct feature of our proposed framework is that the number of sensors attached to end-users is reduced. Additionally, with the use of our framework, the arm orientations can be estimated with any soft sensor, and good accuracy of the estimations can be ensured. Another contribution is the suggestion of the combination of the Bi-LSTM and two-layer LSTM.

This paper studies verification of a combination of hand gestures recognized by using the Myo armband as an input device. To this end, relationship between data distribution and learning curves is investigated for binary classification and multi-class classification problems. A verification method is then proposed for finding a combination of gestures accurately classified. Experiments show effectiveness of the proposed method.

Comunicació presentada a: 5th international Workshop on Sensor-based Activity Recognition and Interaction celebrat el 20 i 21 de setembre de 2018 a Berlin, Alemanya. Motion Capture (MOCAP) Systems have been used to analyze body motion and postures in biomedicine, sports, rehabilitation, and music. With the aim to compare the precision of low-cost devices for motion tracking (e.g. Myo) with the precision of MOCAP systems in the context of music performance, we recorded MOCAP and Myo data of a top professional violinist executing four fundamental bowing techniques (i.e. Détaché, Martelé, Spiccato and Ricochet). Using the recorded data we applied machine learning techniques to train models to classify the four bowing techniques. Despite intrinsic differences between the MOCAP and low-cost data, the Myo-based classifier resulted in slightly higher accuracy than the MOCAP-based classifier. This result shows that it is possible to develop music-gesture learning applications based on low-cost technology which can be used in home environments for self-learning practitioners. This work has been partly sponsored by the Spanish TIN project TIMUL (TIN 2013-48152-C2-2-R), the European Union Horizon 2020 research and innovation programme under grant agreement No. 688269 (TELMI project), and the Spanish Ministry of Economy and Competitiveness under the Maria de Maeztu Units of Excellence Programme (MDM-2015-0502).

Recent years have seen revolutionary changes in CAS field thanks to the introduction of advanced Augmented Reality technologies. In this brief will be presented the overall study carried out in order to develop an AR application to support the doctor during the surgery. It will be described the way the surgeon can perform a more accurate diagnosis by using not just the traditional bidimensional grey-scale CT/MRI images, but also a more realistic three-dimensional marker-based visualization of the patient’s organs. Specific focus will be given to the extremely beneficial way the surgeon can interact with this kind of digital information in the operating room. It will be presented a highly innovative gesture-control device which allows the doctor to interact with IT equipment in a touchless and as natural as possible way, without coming into contact with it or interrupting the surgery, thus preserving the surgical environment from the danger of contamination.

A peppergram allows a user to experience a visual image floating in the air. This paper explores how to add interactive controls to a peppergram. An investigation was conducted to explore the suitability of two different integrative methods for navigating a peppergram as a multimedia presentation tool; a Myo armband, and touch screen input. These methods were used with prototype 3D visual images experienced on a high-resolution mobile tablet device. A pilot study was conducted to evaluate the user experience. We found that using freehand gesture input with the Myo could be one way to provide interaction with peppergram virtual content without requiring any touch input. These results could be used as the basis for further development of interactive peppergram displays. Refereed/Peer-reviewed

This paper presents the development of a Robot Operating System (ROS)-based mobile robot control using electromyography (EMG) signals. The proposed robot&rsquo, s structure is specifically designed to provide modularity and is controlled by a Raspberry Pi 3 running on top of an ROS application and a Teensy microcontroller. The EMG muscle commands are sent to the robot with hand gestures that are captured using a Thalmic Myo Armband and recognized using a k-Nearest Neighbour (k-NN) classifier. The robot&rsquo, s performance is evaluated by navigating it through specific paths while solely controlling it through the EMG signals and using the collision avoidance approach. Thus, this paper aims to expand the research on the topic, introducing a more accurate classification system with a wider set of gestures, hoping to come closer to a usable real-life application.

Technological developments to support the current learning system are so fast that there is an interactive innovation technology for educational trends. One of the technologies implemented is an interactive presentation application in a multimedia class or smart presentation system. This technology makes it possible to control the presentation in a natural way with their hand movements. This introduction can replace conventional mouse roles and functions to facilitate teacher performance in applying interactive technology in the classroom. To build this intelligent presentation system, it is divided into several parts: 1) Recognition sensor arm movement using Myo armband; 2) Hand gesture of hand movements made several steps include: a) data retrieval based on realtime and wireless; b) feature extraction; c) classification using artificial neural network; and 3) Smart presentation, is a presentation system that can understand human behavior and provide interactive presentations.The expected benefits of the results of this study are, with the construction of intelligent presentation systems using hand-gesturing recognition based on the classification of electromyography signals, 1) Make presentations more efficient, engaging and easier to understand, and also make the discussion more interactive and improve communication; 2) Assists the presenter of material in exposing the material by using a presentation control system based on hand gestures.