-A A +A
Psychological health monitoring plays an important role in mood evaluation, especially of ageing subjects within the home environment. For this purpose, the development of innovative and easy-to use platforms based on the use of contact or contactless smart sensor is spreading widely. This paper presents the design and the implementation of a novel framework able to evaluate the mood combining vital signs and facial expressions. For this purpose, a low-cost and commercial vision sensor is used to allow a wider diffusion of the proposed solution and with the aim of increasing the acceptability of the proposed solution. Specifically, a heart rate estimation algorithm and a facial expression recognition module are combined to evaluate the end user’s mood. This result has been achieved through use of deep learning and transfer learning algorithms that work in real time also on embedded hardware platform not equipped with GPUs, consequently increasing its usability. The first added value of the proposed framework consists in the possibility of detecting facial expressions “in the wild” independently from the selected vision sensor and from face orientation. Another important added value lies in the implementation of rule-based expert system which combines data acquired from the same smart sensor but whose operation is also maintained using information from heterogeneous sensors that provide the same type of discrete input values. Due to COVID-19 restrictions, the overall system is currently being tested first in a controlled environment and then in a real environment to achieve the final goal. The findings of the preliminary experiments …
Publication date: 
1 Jan 2021

Andrea Manni, Andrea Caroppo, Pietro Siciliano, Alessandro Leone

Biblio References: 
Pages: 16-25