Toward Machine Emotional Intelligence: Analysis of Affective Physiological State
IEEE Transactions on Pattern Analysis and Machine Intelligence - Graph Algorithms and Computer Vision
| Hi-index | 0.00 |
Emotion recognition is becoming an increasingly important field for human affective computing. This paper presented a method of emotion recognition using Electrocardiography (ECG) signal obtained from multiple subjects. ECG signals were collected when film clips shown to subjects. Through denoising and location of P-QRS-T wave by wavelet transform, ECG features could be extracted effectively. For recognizing two emotions (joy and sadness), Genetic Simulated Annealing Algorithm (GASA) was applied for feature selection. And classification accuracy of fisher classifier was used as evaluation criteria to select the optimal feature subset. Experiment results showed that the research had acquired a good emotion recognition effect and effective emotion features were achieved availably. And it was feasible to recognize the emotion states using GASA and fisher classifier based on ECG signal.