Neuro-fuzzy and soft computing: a computational approach to learning and machine intelligence
Neuro-fuzzy and soft computing: a computational approach to learning and machine intelligence
| Hi-index | 0.00 |
In this paper, we propose an emotion recognition system for understanding the emotional state of human reflected from a movie clip. In order to analyze the human emotion, we consider the electroencephalogram (EEG) signals which are stimulated while watching movie clips to form the semantic emotional dynamic features. These features are then used to analyze the emotional state of human mind stimulated by emotional scene in movie clips. Changes in alpha and gamma power have been interpreted to indicate differential valence patterns related to the frontal lobes. More active left frontal region indicates a positive reaction, and more active right anterior lobe indicates negative affection. So, the alpha and gamma power in the EEG signals are used to obtain EEG features that recognize the emotional states. In order to extract the emotional feature in a movie clip from EEG signals, both independent component analysis (ICA) which rejects the artifact and Short Time Fourier Transform (STFT) are used. Then, we apply the 3-D fuzzy GIST to effectively describe the emotion related EEG signal. The 3-D fuzzy GIST is based on 3-D tensor data consisting of time dependent energy in a specific power band. The obtained 3-D EEG features are used as inputs to an adaptive neuro-fuzzy inference classifier. We use the mean opinion scores as the teaching signals. Experimental results show that the proposed 3-D EEG features can effectively discriminate the positive emotion from the negative ones.