EmotionSense: Emotion Recognition Based on Wearable Wristband

In this paper, we develop an automatic emotion recognition system based on a sensor-enriched wearable wristband. Specifically, in order to obtain physiological data from participants, we first adopt a video induction method which can spontaneously evoke human emotions in a real-life environment. Meanwhile, a questionnaire is designed to record the emotion status of the participants, which can be used as the ground-truth for emotion recognition. Second, we collect multi-modal physiological signals by utilizing three different biosensors (including blood volume pause, electrodermal activity, and skin temperature) embedded in the Empatica E4 wristband. Furthermore, we extract time, frequency and nonlinear features from the collected physiological signals, and adopt the sequence forward floating selection (SFFS) method to search for the best emotion-related features. Finally, we classify different emotions base on SVM using the selected features in aspect of arousal, valence, and four emotions. An overall accuracy of 76% for 15 participants demonstrates that the proposed system can recognize human emotions effectively.