Title :
The Data Location in Emotional Physiological Reaction Experiments: Based on Data Cross-Correlation Analysis
Author :
Pengchao Shangguan ; Guangyuan Liu ; Wanhui Wen
Author_Institution :
Sch. of Electron. & Inf. Eng., Southwest Univ., Chongqing, China
Abstract :
In emotional physiological reaction experiments, to pinpoint the physiological signals which contain reliable emotional physiological responses is significant for subsequent emotion analysis and recognition. Combined with the continuous emotional rating method, this article uses random matrix theory to analyze the cross-correlation of some pulse physiological signals which are collected from more than 100 ordinary university students about happy, anger, fear and sadness four kinds of emotions. By this method, we attain the precise location of these physiological signal segments with reliable emotional responses in complete physiological signals, to ensure the data validity of subsequent affective computing.
Keywords :
behavioural sciences; correlation methods; emotion recognition; matrix algebra; random processes; signal processing; affective computing; continuous emotional rating method; data cross-correlation analysis; data location; emotion analysis; emotion recognition; emotional physiological reaction experiments; emotional physiological response; ordinary university students; physiological signal segment; physiological signals; pulse physiological signal; random matrix theory; Correlation; Covariance matrices; Eigenvalues and eigenfunctions; Emotion recognition; Materials; Physiology; Time series analysis; cross-correlation analysis; data location; emotional physiological reaction experiments; pulse signals;
Conference_Titel :
Computational Intelligence and Design (ISCID), 2013 Sixth International Symposium on
Conference_Location :
Hangzhou
DOI :
10.1109/ISCID.2013.148