One-shot Learning from Demonstration Approach Toward a Reciprocal Sign Language-based HRI.

This paper addresses the lack of proper Learning from Demonstration (LfD) architectures for Sign Language-based Human-Robot Interactions to make them more extensible. The paper proposes and implements a Learning from Demonstration structure for teaching new Iranian Sign Language signs to a teacher assistant social robot, RASA. This LfD architecture utilizes one-shot learning techniques and Convolutional Neural Network to learn to recognize and imitate a sign after seeing its demonstration (using a data glove) just once. Despite using a small, low diversity data set (~ 500 signs in 16 categories), the recognition module reached a promising 4-way accuracy of 70% on the test data and showed good potential for increasing the extensibility of sign vocabulary in sign language-based human-robot interactions. The expansibility and promising results of the one-shot Learning from Demonstration technique in this study are the main achievements of conducting such machine learning algorithms in social Human-Robot Interaction.

Objective measurement of Inferior-Directed stiffness in glenohumeral joint using a specially designed robotic device in healthy shoulders; Within- and Between-Session reliability.

Clinical assessment of capsuloligamentous structures of the glenohumeral joint has been qualitative and subjective in nature, as demonstrated by limited intra- and inter-rater reliability. Robotic devices were utilized to develop a clinically objective measurement technique for glenohumeral joint stiffness. The purpose of this study was to quantify the amount of inferior-direction stiffness of the glenohumeral joint using a safe clinical device in the asymptomatic individuals, and to determine between trial and between session reliability of the robotic device. Twenty healthy subjects were recruited via convenience sampling. Inferior-directed translation and applying force were measured using displacement and force sensors of a robotic device. The stiffness values were calculated as the mean of the slopes of the linear portions of the force-displacement curves for the cycles obtained after familiarization and preconditioning. Four trials for each measurement occasion were averaged to determine the stiffness value for each subject in one session. Repeatability of glenohumeral joint stiffness measurements for between trials and between two sessions was determined using intraclass correlation values and standard error of the measurements. The mean stiffness value was 1.50 N/mm (±0.40) and 1.52 N/mm (±0.40), respectively. The robotic device for stiffness assessment was reliable for repeated measures of stiffness in one session, and between sessions with ICC equal 0.96 (95% CI 0.93-0.98), and 0.97 (95% CI 0.95-0.99), respectively. The SEM between the trials was in each session 0.08 N/mm. The results of this study provide that our robotic technique for quantifying glenohumeral joint stiffness is precise and reproducible.