Emotion recognition based on convolutional neural networks and heterogeneous bio-signal data sources

Wang Kay Ngai, Haoran Xie, Di Zou, Kee-Lee Chou

Abstract: Highlights•A multi-branch convolutional neural network integrates heterogenous data sources.•Data sources include EEG data, face data and eye data for emotion recognition.•SAM questionnaires with valence and arousal levels are used for emotion labeling.•The proposed emotion recognition outperforms existing comparable baselines.