An Emotion Recognition Approach based on Wavelet Transform and Second-Order Difference Plot of ECG

Emotion, as a psychophysiological state, plays an important role in the human communications and daily life. Emotion studies related to the physiological signals have recently been the subject of many research works. In this work, a hybrid feature-based approach is proposed to examine the affective states. To this effect, the electrocardiogram (ECG) signals of 47 students are recorded using the pictorial emotion elicitation paradigm. Affective pictures are selected from the International Affective Picture System and assigned to four different emotion classes. After extracting the approximate and detailed coefficients of Wavelet Transform (WT/Daubechies 4 at level 8), two measures of the second-order difference plot (CTM and D) are calculated for each wavelet coefficient. Subsequently, Least Squares Support Vector Machine (LS-SVM) is applied to discriminate between the affective states and the rest. The statistical analysis results indicate that the CTM density in the rest is distinctive from the emotional categories. In addition, the second-order difference plot measurements at the last level of WT coefficients show significant differences between the rest and the emotion categories. Applying LS-SVM, a maximum classification rate of 80.24% was reached for discrimination between the rest and the fear. The results of this study indicate the usefulness of WT in combination with the non-linear technique in characterizing the emotional states.