Emotion Identification With Smartphones to Improve the Elder Quality of Life Using Facial Recognition Techniques

Emotion Identification With Smartphones to Improve the Elder Quality of Life Using Facial Recognition Techniques

Sheila Bonilla, Enrique Moguel, José Garcia-Alonso, Javier Berrocal, Juan M. Murillo
Copyright: © 2020 |Pages: 16
DOI: 10.4018/978-1-7998-1937-0.ch011
(Individual Chapters)
No Current Special Offers


The increase in the elderly population today is a fact. This group of people needs day-to-day care due to their age, and, in addition, they often have health problems. Technology can be used to mitigate these problems. However, it must be borne in mind that most of this population is currently unable to get the most out of electronic devices. To help elders benefit from these devices, systems adapted to their needs, and preferences are needed. In particular, systems that use the elders' contextual information to integrate several aspects of eldercare and adapt them to each elder would provide significant benefits. In this case, the emotions will be used to recognize to what extent an elderly person needs care at certain times of the day and to adapt surrounding IoT systems to their needs and moods. For this purpose, this chapter proposes to use smartphones as the devices that centralize contextual information of the elders, focusing on emotion recognition.
Chapter Preview


Today, thanks to advances in technology and communications, one of the fastest advancing electronic devices is the mobile phone. Nowadays, a large part of the population uses their mobile phones to communicate, search information, socialize, etc. Even in the context of elder population, there has been an increase in the uses of smartphones in recent years (Sánchez López, Fernández Alemán, Toval, & Carrillo de Gea, 2015).

A large part of the population in developed countries is formed by elder people (65 years of older). It is estimated that by 2030 this age group may reach 26% of the population (Rodrigues, Huber, Lamura, et al., 2012). On the other hand, in Spain, since 2012, 75.9% of people over the age of 65 use a mobile phone and, in addition, the use of this device has become a daily habit for them (Sánchez López et al., 2015).

In order to help this sector of the population, in addition to using mobile phones, more IoT (Internet of Things) devices are used everyday. These devices are connected to everyday objects through the Internet to add new or improved functionalities to them. Frequently, the elderly tend to use this type of device without being aware of it. For example, through medical care in intelligent homes, where the main objective of these systems is to have better control over the patient’s health care, significantly reducing hospital visits and improving their quality of life. Another example in which IoT devices are used is the intelligent medication service, since the success in the treatment of any person depends to a great extent on taking the prescribed medicines at the right time. Due to this, smart pillbox or even smart pharmaceutical packaging were created (Núñez, Mendoza, Hernández, & Molinares, 2016) or (Morales Suárez-Varela, 2009). To use these devices need technical skills, since the must be initially con- figured so that they know how to satisfy their owners needs, for example the time at which the elderly should take their medicines. However, if these needs change, the devices must be reprogrammed.

Having this into account, the emotional state of the users is usually over- looked when gathering contextual information. However, the emotions displayed by user when performing certain action are of great interest to improve the systems behavior.

According to (Kulkarni, Shendge, Varma, & Kimmatkar, 2018) there are six basic types of emotions, which are: joy, sadness, surprise, fear, disgust and anger. Being able to identify these emotions will help improve the contextual information gathered in the system, and therefore, obtain systems better adapted to the needs and moods of their users. In the context of the elder population, this technique could be used to perform actions like detecting if the person has a problem at a certain moment of the day or even quite the opposite, due to the fact that this person has a positive emotional state.

For emotion recognition, numerous techniques are currently used, such as facial recognition through images or videos such as (Suk & Prabhakaran, 2014), (Claudino, de Lima, de Assis, & Torro, 2019) or (Gan, Chen, & Xu, 2019), but as far as the authors know it has not been studied how to include this in the everyday life of elderly people. In this book chapter, the authors present a technique to analyse the emotional state of the elderly, through facial recognition of emotions and the subsequent use of this information as part of the Situational Context, which allows for a better adaptation of IoT systems to the emotional state of elderly people.

In order to present our proposal, the rest of chapter is structured as follows: In Section 2, we will detail the background of this work focusing on emotion recognition and adaptation techniques based on contextual information; in Section 3, we will describe our proposal in which we will highlight the architecture and two case studies. Finally, in the last Section we will show the conclusions and future works.

Complete Chapter List

Search this Book: