Emotion Recognition in Valence-Arousal Scale by Using Physiological Signals

Akalin N., Köse H.

26th IEEE Signal Processing and Communications Applications Conference (SIU), İzmir, Turkey, 2 - 05 May 2018 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume:
  • Doi Number: 10.1109/siu.2018.8404632
  • City: İzmir
  • Country: Turkey
  • Istanbul Technical University Affiliated: Yes

Abstract

Physiological signals are generated autonomously by central nervous system therefore they cannot be controlled or masked. In the literature, physiological signals are commonly used for emotion recognition. In this study, emotional stimulation is induced using visual and audio stimuli. We used Microsoft Band 2 arm band and Mind Wave EEG sensor for recording the participants' signals during the experiment. We collected data from 20 participants and recognized their emotions on the valence-arousal scale. We tested and compared 3 different classifiers for recognition.