DocumentCode
3585557
Title
MAPD: A Multi-subject Affective Physiological Database
Author
Wenjin Huang ; Guangyuan Liu ; Wanhui Wen
Author_Institution
Coll. of Electron. & Inf. Eng., Southwest Univ., Chongqing, China
Volume
2
fYear
2014
Firstpage
585
Lastpage
589
Abstract
A multi-subject affective physiological database containing 380 physiological records of 250 subjects is presented in this paper. While the subjects individually watch the amusement, anger, fear, or sadness elicitation film clip, their oxygen saturation (OXY), galvanic skin response (GSR), electrocardiogram (ECG), and front face videos are synchronously recorded. The subjects have reported the duration and category of their emotion experience by means of pressing a button during the experiment and filling out a revised PANAS-X after the experiment, respectively. A statistical analysis of the subjects´ emotion experience self-reports is carried out to verify the effectiveness of the emotion elicitation. In order to explore the distinguish ability of the affective physiological changes of the above-mentioned four emotions, 233 affective physiological features commonly used in literature are extracted to represent the 380 affective physiological data samples. A random forests classifier is applied for quaternary classification of amusement, anger, fear, and sadness, and the overall correct rate is 74.8%.
Keywords
learning (artificial intelligence); medical computing; pattern classification; physiology; statistical analysis; video databases; ECG; GSR; MAPD database; OXY; PANAS-X; affective physiological change; amusement; anger; electrocardiogram; emotion elicitation; emotion experience; fear; front face video; galvanic skin response; multisubject affective physiological database; oxygen saturation; random forests classifier; sadness; statistical analysis; Affective computing; Databases; Electrocardiography; Feature extraction; Films; Physiology; Videos; Affective Computing; Affective Physiological Signal; Emotion Recognition;
fLanguage
English
Publisher
ieee
Conference_Titel
Computational Intelligence and Design (ISCID), 2014 Seventh International Symposium on
Print_ISBN
978-1-4799-7004-9
Type
conf
DOI
10.1109/ISCID.2014.247
Filename
7082059
Link To Document