Psychological Internal State Estimation based on Expressive Tempos and Rhythms of Facial Expressions

2017 Fiscal Year Final Research Report

• Project/Area Number: 25330325
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Kansei informatics
• Research Institution: Akita Prefectural University
• Principal Investigator: Sato Kazuhito 秋田県立大学, システム科学技術学部, 教授 (80390904)
• Research Collaborator: Kadowaki Sakura
• Project Period (FY): 2013-04-01 – 2018-03-31
• Keywords: 心理的ストレス / 顔表情 / 機械学習 / グラフィカルモデル / 情動喚起ビデオ / 表出テンポ / 表出リズム

Outline of Final Research Achievements

In this study, ｗe tried to develop a stress model graphically expressing the causal relationship between psychological stress levels on daily basis and facial expressions. As a result, we revealed that facial expressions easily affected by stress levels are "happiness" and "sadness" facial expressions, and facial parts easily appearing, such as eyes and mouth, differ depending on the type of facial expressions. Furthermore, particularly addressing the expressive process of "happiness" facial expression after giving a pleasant-unpleasant stress stimulus and the degree of mutual coordination among facial parts, "Unnatural smile" and "Natural smile" can be distinguished, and the analysis of facial expressions showed high practicality as a window of mental health for human.

Free Research Field

システム工学