EMOTIONS RECOGNITION BASED ON EEG TIME DOMAIN FEATURE EXTRACTION
Ma Li, Abdul Wahab Abdul Rahman and KaiXiang Teo
International Symposium on Performance Evaluation of Computer and Telecommunication Systems (SPECTS 2007)
San Diego, California (USA), July 16-18, 2007
SPECTS_Summary
In this paper we propose an emotion recognition system using the EEG signal based on time domain analysis for extracting the emotion features. Three different types of emotions video clip, mainly: happy, relax and sad, are used to prime the subjects. Subjective surveys are first carried out prior to monitoring the EEG signals. Experimental results based on SVM, RVM, DT, MLP and Bayesian show the potential of using the EEG time domain analysis for real-time application.