We propose a sign language recognition system based on wearable electronics and two different classification algorithms. The wearable electronics were made of a sensory glove and inertial measurement units to gather fingers, wrist, and arm/forearm movements. The classifiers were k-Nearest Neighbors with Dynamic Time Warping (that is a non-parametric method) and Convolutional Neural Networks (that is a parametric method). Ten sign-words were considered from the Italian Sign Language: cose, grazie, maestra, together with words with international meaning such as google, internet, jogging, pizza, television, twitter, and ciao. The signs were repeated one-hundred times each by seven people, five male and two females, aged 29–54 y ± 10.34 (SD). The adopted classifiers performed with an accuracy of 96.6% ± 3.4 (SD) for the k-Nearest Neighbors plus the Dynamic Time Warping and of 98.0% ± 2.0 (SD) for the Convolutional Neural Networks. Our system was made of wearable electronics among the most complete ones, and the classifiers top performed in comparison with other relevant works reported in the literature.

We investigate Relational Graph Attention Networks, a class of models that extends non-relational graph attention mechanisms to incorporate relational information, opening up these methods to a wider variety of problems. A thorough evaluation of these models is performed, and comparisons are made against established benchmarks. To provide a meaningful comparison, we retrain Relational Graph Convolutional Networks, the spectral counterpart of Relational Graph Attention Networks, and evaluate them under the same conditions. We find that Relational Graph Attention Networks perform worse than anticipated, although some configurations are marginally beneficial for modelling molecular properties. We provide insights as to why this may be, and suggest both modifications to evaluation strategies, as well as directions to investigate for future work.

Normal communication of deaf people in ordinary life still remains an unrealized task, despite the fact that Sign Language Recognition (SLR) made a big improvement in recent years. We want here to address this problem proposing a portable and low cost system, which demonstrated to be effective for translating gestures into written or spoken sentences. This system relies on a home-made sensory glove, used to measure the hand gestures, and on Wavelet Analysis (WA) and a Support Vector Machine (SVM) to classify the hand's movements. In particular we devoted our efforts to translating the Italian Sign Language (LIS, Linguaggio Italiano dei Segni), applying WA for feature extractions and SVM for the classification of one hundred different dynamic gestures. The proposed system is light, not intrusive or obtrusive, to be easily utilized by deaf people in everyday life, and it has demonstrated valid results in terms of signs/words conversion.

Hands gestures recognition, by means of measuring apparatus, can provide a new way of human-computer interaction. Controlling different devices or speaking through a speech synthesizer can be time saving as well as an aid for impaired persons. In this work we performed the classification of 20 different gestures, evaluating three different methodologies: Support Vector Machines, Mahalanobis and Euclidean based classifiers.

This paper carried out a statistical analysis of human finger’s joint angles during hand specific daily activities, studying the correlations among the joints and applying a linear regression to express their correlations. The aim was to reduce the number of myoelectric sensors necessary in devices such as prosthesis, stands the current surgery difficulties and the problem of rejection, but without losing too many degrees of freedom. Measures were taken using our special hand movement acquisition system called HITEG data glove. As a preliminary work, we decided to limit the set of gestures performed to 9 of the most common movements of the human hand. The results shown that the number of sensors can be reduced from 14 to 7 with an acceptable error on the presumed value of each finger joint angle which can be as low as 10 degrees.

The understanding of surgical gesture, by means of measuring apparatus, can play a key role for a possible evaluation of the surgical performance and the human factors characterizing it. To this aim a neural network classification algorithm can be helpful, since combines good generalization performances along with a parsimonious architecture when dealing with high dimensional classification problems. So, here it is presented and proposed the development of an innovation in surgical training system, as a fundamental objective support for training of novice surgeons.

A Support Vector Machine (SVM) classification method for data acquired by EEG registration for brain/computer interface systems is here proposed. The aim of this work is to evaluate the SVM performances in the recognition of a human mental task, among others. Such methodology could be very useful in important applications for disabled people. A prerequisite has been the developing of a system capable to recognize and classify the following four tasks: thinking to move the right hand, thinking to move the left hand, performing a simple mathematical operation, and thinking to a nursery rhyme. The data set exploited in the training and testing phases has been acquired by means of 61 EEG electrodes and consists of several time series. These time data sets were then transformed into the frequency domain, in order to obtain the power frequency spectrum. In such a way, for every electrode, 128 frequency channels were obtained. Finally, the SVM algorithm was used and evaluated to get the proposed classification.

In this study a comparison among three different machine learning techniques for the classification of mental tasks for a Brain-Computer Interface system is presented: MLP neural network, Fuzzy C-Means Analysis and Support Vector Machine (SVM). In BCI literature, finding the best classifier is a very hard problem to solve, and it is still an open question. We considered only ten electrodes for our analysis, in order to lower the computational workload. Different parameters were analyzed for the evaluation of the performances of the classifiers: accuracy, training time and size of the training dataset. Results demonstrated how the accuracies of the three classifiers are nearly the same but the error margin of SVM on this reduced dataset is larger compared to the other two classifiers. Furthermore neural network needs a reduced number of trials for training purposes, reducing the recording session up to 8 times with respect to SVM and Fuzzy analysis. This suggests how, in the presented case, MLP neural network can be preferable for the classification of mental tasks in Brain Computer Interface systems.

This study is devoted to the classification of four-class mental tasks data for a brain-computer interface protocol. In such view we adopted multi layer perceptron neural network (MLP) and fuzzy C-means analysis for classifying: left and right hand movement imagination, mental subtraction operation and mental recitation of a nursery rhyme. Five subjects participated to the experiment in two sessions recorded in distinct days. Different parameters were considered for the evaluation of the performances of the two classifiers: accuracy, that is, percentage of correct classifications, training time and size of the training dataset. The results show that even if the accuracies of the two classifiers are quite similar, the MLP classifier needs a smaller training set to reach them with respect to the fuzzy one. This leads to the preference of MLP for the classification of mental tasks in brain computer interface protocols.

Body Area Network (BAN) and Wireless BAN (WBAN) systems lack a unique description model for the identification of all the components and the features that characterize them. This results in a disadvantage since a formalization tool would favor standardization and would seriously help in the implementation of building systems. In this paper we successfully used the Unified Modeling Language (UML) to describe a WBAN which measures hand joint movements, and we demonstrated how UML can be successfully adopted to design a model for the description of such networks.