Abstract
New and more natural human-robot interfaces are of crucial interest to the evolution of robotics. This paper addresses continuous and real-time hand gesture spotting, i.e., gesture segmentation plus gesture recognition. Gesture patterns are recognized by using artificial neural networks (ANNs) specifically adapted to the process of controlling an industrial robot. Since in continuous gesture recognition the communicative gestures appear intermittently with the non-communicative, we are proposing a new architecture with two ANNs in series to recognize both kinds of gesture. A data glove is used as interface technology. Experimental results demonstrated that the proposed solution presents high recognition rates (over 99% for a library of ten gestures and over 96% for a library of thirty gestures), low training and learning time and a good capacity to generalize from particular situations.

Abstract

Abstract

Abstract
The aim of this paper was to identify active drag determinants and classify swimmers based on such features. 67 young swimmers made a maximal 25m Front-Crawl to measure with a speedo-meter the swimming velocity (v), speed-fluctuation (dv) and dv normalized to v (dv/v). Another two 25m bouts with and without a perturbation device were made to estimate active drag coefficient (CDa). Trunk transverse surface area (S) was measured with photogrammetric technique on-land and in the hydrodynamic position. Cluster 1 was related to swimmers with a high speed fluctuation (i.e., dv and dv/v). Cluster 2 was characterized by the anthropometrics (i.e., S). Cluster 3 was associated with the high hydrodynamic profile (i.e.,CDa). The variable that seems to discriminate better the clusters was the dv/v (F=53.680; P<0.001), followed by the dv (F=28.506; P<0.001), CDa (F=21.025; P<0.001), S (F=6.297; P<0.01) and v (F=5.375; P=0.01). Stepwise discriminant analysis extracted 2 functions. Function 1 was mainly defined by dv/v and S (74.3% of variance), while Function 2 was mainly defined by CDa (25.7% of variance). So, it can be concluded that kinematics, anthropometrics and hydrodynamic features are determinant domains to classify and characterize swimmers’ profiles.

Abstract
Assistive Technologies may greatly contribute to give autonomy and independence for individuals with physical limitations. Electric wheelchairs are examples of those assistive technologies and nowadays each time becoming more intelligent due to the use of technology that provides assisted safer driving. Usually, the user controls the electric wheelchair with a conventional analog joystick. However, this implies the need for an appropriate methodology to map the position of the joystick handle, in a Cartesian coordinate system, to the wheelchair wheels intended velocities. This mapping is very important since it will determine the response behavior of the wheelchair to the user manual control. This paper describes the implementation of several joystick mappings in an intelligent wheelchair (IW) prototype. Experiments were performed in a realistic simulator using cerebral palsy users with distinct driving abilities. The users had 6 different joystick control mapping methods and for each user the usability and the users' preference order was measured. The results achieved show that a linear mapping, with appropriate parameters, between the joystick's coordinates and the wheelchair wheel speeds is preferred by the majority of the users.

Abstract