Utilization of Real-Time Heart Rate Changes to Classify Emotions Based on Low-Cost Device Development

• Main Authors: Sofyan, Pramudya Rakhmadyansyah, Wahyudi, Rizdha, Putra, Diandri Perkasa, Sahroni, Alvin, Widiasmara, Nur, Miladiyah, Isnatin
• Format: Conference Proceeding
• Language: English
• Subjects: Education; Emotion recognition; Emotions; Feature extraction; Finite impulse response filters; Heart rate; Heart rate changes; Low-cost device; Machine learning; Real-time systems
• ISSN: 2189-8723
• DOI: 10.1109/ICIIBMS52876.2021.9651613

Emotion is a complex phenomenon and has an essential role in supporting the quality of life and affecting motivation, perception, cognition, creativity, empathy, education, and making an important decision. Also, negative emotions such as anger, shame, and anxiety usually arise from stress, and it is referring to destructive and threatening to show a relationship between them. Therefore, research on recognizing emotions is still an important issue currently. Fifteen males university students have participated in this study. A fingertip photoplethysmograph (PPG) was used to record the heart rate variable. This study used The International Affective Picture System (IAPS) to induce particular emotions. As a subjective evaluation, we employed Self-Assessment Manikin (SAM) to evaluate the subject's emotions. Then, we compared the SAM evaluation to their heart rate properties for further investigation. The FIR Band Pass Filter was established as a pre-processing method, and a single feature was extracted from a PPG recording called Heart Rate Changes (HCR). Instead of utilizing artificial intelligence or other machine learning classification techniques, we performed if-else classification to recognize negative and positive emotions as basic emotions. We found that the HCR was significantly different between the positive and negative emotions (p