HUMAN EMOTION INTENSITY ESTIMATION: A SURVEY OF CATEGORIZATION AND TOWARDS THERMAL IMAGERY

Abstract

Emotion is an important factor for social interaction. In the era of the explosion of  artificial intelligence, developing a system that can automatically estimate  human emotions has attracted enormous interest to researchers. Understanding inner emotion makes human-computer interaction applications be able to join mankind life extensively. Most of human emotion estimation works  focus on categorizing emotion based on visible images or videos of facial expression, but only few works address the use of infrared thermal imaging for emotion estimation, which has significant advantages in the field. In additional, we still lack a good survey of human emotion estimation using infrared thermal imaging. Therefore, in this paper, we concentratedly survey about human emotion estimation exploiting both visible and infrared thermal imaging. Firstly, we want to give a review of the most  recent studies about human emotion estimation, which include emotion categorization and emotion intensity  estimation. Secondly, we identify several challenges in this field and suggest some idea for future research.

Keywords: Facial expression, emotion categorization, emotion estimation, emotion intensity, visible image, thermal image.