Emotional Stress Recognition using Multi-Modal Bio-Signals

Human emotional assessment helps in identification of human stress, which is detected using multi-modal bio-signals. Multi-modal bio-signal involves EEG signals and psycho-physiological signals such as Skin Conductance (SC), Blood Pressure (BP), Heart Rate Variability, and Respiration. The raw EEG signal and psycho-physiological signals were pre-processed and decomposed into five different frequency bands (delta, theta, alpha, beta, gamma) using Discrete Wavelet Transform (DWT). In this work, we used two different wavelet functions namely db8 and sym8 for extracting the statistical features from EEG signal and psycho-physiological signals for classifying the emotional stress. In order to evaluate the efficiency of emotional stress, Support Vector Machine (SVM) is used.