Consideration of Appropriate Photo Image to the Conversation of the Elderly and Youth by Using Facial Expression Analysis

Abstract

Conversation is a good preventative against behavioral problems in the elderly. However, caregivers are usually very busy tending to patients and lack the time to communicate extensively with them. In order to overcome such problems active “listening volunteers” have more opportunities to communicate with the elderly, but the number of skillful volunteers is limited. Therefore, we investigated conversational support systems which can help inexperienced volunteers; such systems typically have included content such as photographs, videos, and music, and in our case we primarily used photos. We predicted that whether, and how much, frustration and discomfort the young adult volunteers experienced through such conversations depended on the category of the photo. We measured dissatisfaction and discomfort during conversations by using an expression analysis sensor. Furthermore, we took the data obtained by measuring dissatisfaction and discomfort through obtained from non-contact devices (expression analysis) as an object of comparison and evaluation.