Establishment of the Chinese facial emotion images database with intensity classification
10.3969/j.issn.1000-6729.2019.02.009
- VernacularTitle:中国人面孔表情强度分级图片库的建立
- Author:
Jing LIU
1
;
Yanping REN
;
Wei QU
;
Nan CHEN
;
Hongzhen FAN
;
Lili SONG
;
Meng ZHANG
;
Zhanxiao TIAN
;
Yanli ZHAO
;
Shuping TAN
Author Information
1. 北京回龙观医院
- Keywords:
emotion;
facial expression;
emotional stimulus material;
assessment
- From:
Chinese Mental Health Journal
2019;33(2):120-125
- CountryChina
- Language:Chinese
-
Abstract:
Objective: To establish an intensity-classified Chinese facial emotion image database and provide materials for emotional researches in China. Methods: Firstly, under the guidance of professional director, 22 professional dramatic actors put up seven kinds of emotion, including anger, disgust, fear, sadness, surprise, happy and calm. The process of facial emotional performance of actors were filmed by a high-speed camera. Then we made screen captures according to the intensity of emotion performed by the actors. A total of 814 black-and-white photographs with same size and gray level were made after we removed the irrelevant parts to the facial features. Each kind of emotion contains six levels of intensity scoring from 1 to 6. Totally, 112 healthy volunteers were recruited to evaluate the emotional intensity, pleasure and arousal of all images via a visual scale ranged from 0 to 100. Results: A Chinese face emotional intensity hierarchical picture library was set up, which contained 2 sublibraries. The first library was consisted of 436 images with hierarchical emotional intensity images collected from different performers. The second one was built with 640 images, in which the hierarchical emotional intensity image was collected from same actor/actress. Both image libraries had corresponding scores for intensity, pleasure, and arousal. Furthermore, both exhibited a high degree of internal consistency in reliability analysis (with a Cronbach' s alpha value larger than 0. 9). Conclusion: The present study has established an intensity-classified Chinese facial emotion image database with good recognition degree and psychological measurement attribution, which could provide materials for emotional researches in China.
