This paper proposes and evaluates the application of support vector machine (SVM) to classify upper limb motions using myoelectric signals. It explores the optimum configuration of SVM-based myoelectric control, by suggesting an advantageous data segmentation technique, feature set, model selection approach for SVM, and postprocessing methods. This work presents a method to adjust SVM parameters before classification, and examines overlapped segmentation and majority voting as two techniques to improve controller performance. A SVM, as the core of classification in myoelectric control, is compared with two commonly used classifiers: linear discriminant analysis (LDA) and multilayer perceptron (MLP) neural networks. It demonstrates exceptional accuracy, robust performance, and low computational load. The entropy of the output of the classifier is also examined as an online index to evaluate the correctness of classification; this can be used by online training for long-term myoelectric control operations.

The goal of our research is to develop socially acceptable behavior for domestic robots in a setting where a user and the robot are sharing the same physical space and interact with each other in close proximity. Specifically, our research focuses on approach distances and directions in the context of a robot handing over an object to a user. Our present study consisted of two parts. Firstly, we carried out a large-scale survey trying to identify the internal structure between users’ different social role expectations and the relationships between these and users’ technology usage. Results from this study led to the development of a measure for preferred robot social roles. In the second part, this measure was used in a live human–robot interaction (HRI) study in a home setting, designed to create a baseline understanding of human–robot Proxemics approach directions and distances for a Care-O-bot \(^{\circledR }\) 3 based on these roles. In order to support users interpretation of the robots intentions in those tasks, the HRI studies used a simple LED light display panel. Findings indicate that, participants were comfortable with the robot approaching to the closest implemented distance (0.5 m). For the task requiring relatively more coordination, participants preferred the robot to approach from the front to a larger extent than for tasks requiring less coordination. The ability to identify the signals from the LED display also impacted how participants evaluated the robots behaviour. Users who had previous experience of interacting with robots differentiated less between approaches, and also evaluated tasks requiring more coordination with the robot more favourably. The findings in the short-term sample were strongly influenced by expectations as to the social role expectation they had of the robot, suggesting that social expectations impact proxemic interactions even if the robot platform is clearly not humanoid.

A long-term Human-Robot Proxemic (HRP) study was performed using a newly developed Autonomous Proxemic System (APS) for a robot to measure and control the approach distances to the human participants. The main findings were that most HRP adaptation occurred in the first two interaction sessions, and for the remaining four weeks, approach distance preferences remained relatively steady, apart from some short periods of increased distances for some participants. There were indications that these were associated with episodes where the robot malfunctioned, so this raises the possibility of users trust in the robot affecting HRP distance. The study also found that approach distances for humans approaching the robot and the robot approaching the human were comparable, though there were indications that humans preferred to approach the robot more closely than they allowed the robot to approach them in a physically restricted area. Two participants left the study prematurely, stating they were bored with the repetitive experimental procedures. This highlights issues related to the often incompatible demands of keeping experimental controlled conditions vs. having realistic, engaging and varied HRI trial scenarios.

This paper presents an ongoing investigation to select optimal subset of features from set of well-known myoelectric signals (MES) features in time and frequency domains. Four channel of myoelectric signal from upper limb muscles are used in this paper to classify six distinctive activities. Cascaded genetic algorithm (GA) has been adopted as the search strategy in feature subset selection. Davies-Bouldin index (DBI) and Fishers linear discriminant index (FLDI) are employed as the filter objective functions and linear discriminant analysis (LDA) has been used as the wrapper objective function. Results prove more accurate and reliable classification for the elite subset of features applying to artificial neural networks as the classifier.

This paper evaluates supervised and unsupervised adaptive schemes applied to online support vector machine (SVM) that classifies BCI data. Online SVM processes fresh samples as they come and update existing support vectors without referring to pervious samples. It is shown that the performance of online SVM is similar to that of the standard SVM, and both supervised and unsupervised schemes improve the classification hit rate.