Decoding emotional experiences through physiological signal processing

摘要: All modern emotion theoretical views assume a role for peripheral physiological changes during emotional experiences. In this paper, we explored the correlation between autonomically-mediated in multimodal bodily signals and discrete states. order to fully exploit information each modality, week learners based on individual signal modalities are built then fused formed robust inference model. To validate our model, three specific including Electromyogram (EMG), Blood Volume Pressure (BVP) Galvanic Skin Response (GSR) recorded eight states were analyzed. Our approach showed 88.1% recognition accuracy, which outperformed conventional Support Vector Machine (SVM) classifier with 17% accuracy improvement. Furthermore, avoid redundancy resultant over-fitting, feature reduction method is proposed analysis optimize number of features required training validating weak learner. Despite space dimensionality from 27 18 features, methodology preserved about 85.0%.