EEG affect analysis based on KDE and MFCC
Classifying emotions based on the affective states of valence and arousal captured from brain discharge remains a challenge. The selection of the most efficient and reliable method of feature extraction forms a very important problem of EEG signal classification. Different methods applied are usuall...
| Main Authors: | , , , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English |
| Published: |
2012
|
| Subjects: | |
| Online Access: | http://irep.iium.edu.my/32132/ http://irep.iium.edu.my/32132/ http://irep.iium.edu.my/32132/1/32132.pdf |
| Summary: | Classifying emotions based on the affective states of valence and arousal captured from brain discharge remains a challenge. The selection of the most efficient and reliable method of feature extraction forms a very important problem of EEG signal classification. Different methods applied are usually based upon the time-domain or frequency-domain analysis. The following study is devoted to the EEG affect analysis based on feature extraction using KDE and MFCC and comparison of results achieved. MLP is used for classification of the features extracted. The resultant feature vectors extracted using KDE provides a more accurate capture of basic emotions when compared with MFCC feature vectors. |
|---|