Luca Romeo

Evaluation of Unimodal and Multimodal Communication Cues for Attracting Attention in Human–Robot Interaction

One of the most common tasks of a robot companion in the home is communication. In order to initiate an information exchange with its human partner, the robot needs to attract the attention of the human. This paper presents results of two user studies (

Accuracy evaluation of the Kinect v2 sensor during dynamic movements in a rehabilitation scenario

\mathrm{N}=12

A novel computer vision based e-rehabilitation system: From gaming to therapy support

N=12) to evaluate the effectiveness of unimodal and multimodal communication cues for attracting attention. Results showed that unimodal communication cues which involve sound generate the fastest reaction times. Contrary to expectations, multimodal communication cues resulted in longer reaction times with respect to the unimodal communication cue that produced the shortest reaction time.

Modular design of a novel wireless sensor node for smart environments

In this work we propose a real-time detection of mental stress during different cognitive tasks. Stress is classified processing Galvanic Skin Response (GSR), RR Interval and Body Temperature (BT) acquired by a commercial smartwatch. The unobtrusive system proposed is validated through clinical psychological tests.

An instrumental approach for monitoring physical exercises in a visual markerless scenario: A proof of concept

In this paper, the accuracy evaluation of the Kinect v2 sensor is investigated in a rehabilitation scenario. The accuracy analysis is provided in terms of joint positions and angles during dynamic postures used in low-back pain rehabilitation. Although other studies have focused on the validation of the accuracy in terms of joint angles and positions, they present results only considering static postures whereas the rehabilitation exercise monitoring involves to consider dynamic movements with a wide range of motion and issues related to the joints tracking. In this work, joint positions and angles represent clinical features, chosen by medical staff, used to evaluate the subjects movements. The spatial and temporal accuracy is investigated with respect to the gold standard, represented by a stereophotogrammetric system, characterized by 6 infrared cameras. The results provide salient information for evaluating the reliability of Kinect v2 sensor for dynamic postures.

Serious gaming approach for physical activity monitoring: A visual feedback based on quantitative evaluation

This work deals with the design of an interactive monitoring tool for home-based physical rehabilitation. The software platform includes a video processing stage and the exercise performance evaluation. Image features are extracted by a Kinect v2 sensor and elaborated to return the exercises score. Furthermore the tool provides to physiotherapists a quantitative exercise evaluation of subjects performances. The proposed tool for home rehabilitation has been tested on 5 subjects and 5 different exercises and results are presented. In particular both exercises and relative evaluation indexes were selected by specialists in neurorehabilitation.

A Smart Sensing Architecture for Domestic Monitoring: Methodological Approach and Experimental Validation

We propose a novel e-rehabilitation system based on a commercial RGB-D device. Differently from exergaming approaches, clinical objectives scores of each specific body part involved in the exercise are computed. Subjects performances are sent to the physiotherapists in order to support and improve decisions and therapies.

Person Re-Identification with RGB-D Camera in Top-View Configuration through Multiple Nearest Neighbor Classifiers and Neighborhood Component Features Selection

Most of the existing commercial node architectures provide little flexibility and configurability. This limitation constrains the usability of the same node across various applications, including the ambient intelligence issue. In this paper a novel architecture for the design of a modular wireless sensor node is proposed, dividing the connection and sensing functions in two separate boards. The division of the wireless transducer interface module (WTIM) in two independent boards allows to perform in a separate way the connection and sensor interfacing function of the WTIM always respecting IEEE 1451 standards. The versatility of the novel architecture has been tested in two different application scenarios. In the first application the modular node has been used in a factory to monitor the efficiency and reliability of the production line. The designed node has been experimentally tested and results shown. Concerning the second application, a smart home approach is proposed. Using different sensing boards, an architecture to monitor in a non-invasive way several home parameters has been presented.