Abstract
This article studies whether heart sound signals can be used for emotion recognition. First, we built a small emotion heart sound database, and simultaneously recorded the participants’ ECG for comparative analysis. Second, according to the characteristics of the heart sound signals, two emotion evaluation indicators were proposed: HRV of heart sounds (difference between successive heartbeats) and DSV of heart sounds (the ratio of diastolic to systolic duration variability). Then, we extracted linear and nonlinear features from two emotion evaluation indicators to recognize four kinds of emotions. Moreover, we used valence dimension, arousal dimension and valence-arousal synthesis as evaluation standards. The experimental results demonstrated that heart sound signals can be used for emotion recognition. It was more effective to achieve recognition results by combining the features of HRV and DSV of heart sounds. Finally, the average accuracy of four emotion recognitions on valence dimension, arousal dimension and valence-arousal synthesis was up to 96.875%, 88.5417% and 81.25%, respectively.
Highlights
This article studies whether heart sound signals can be used for emotion recognition
(2) A heart sound preselecting and segmentation algorithm based on the template selection was used to automatically calculate HS heart rate variability (HRV) and heart sound DSV (HS DSV), using the following formula: Segmentation results of heart sound signals were recorded in two arrays, and the array length was 2N (N is the number of heart sound cycles)
In high emotional arousal dimensions, the SD12 is lower; in low emotional arousal dimensions, the SD12 is higher. This result showed that emotions of the subjects were fully stimulated when heart sound signals and ECG signals were recorded, proving that heart sound HRV (HS HRV) and HS DSV are effective emotion evaluation indicators and that heart sounds can be used for emotion recognition
Summary
This article studies whether heart sound signals can be used for emotion recognition. Jang et al.[6] showed that the differences in physiological responses among emotions were significant for heart rate (HR), skin conductance level (SCL) and skin conductance response (SCR) These physiological signals, especially ECG effectively reflect the relationship between the heart beating and emotion changes. Heart sound signals and ECG signals are different manifestations of cardiac activity Both can effectively reflect the beating of the heart and changes in emotion[9]. The HS HRV (heart rate variability based on heart sound signals) is a minor difference between successive intervals of heart sound signals. Extracting linear and nonlinear features from HS HRV and HS DSV achieved the emotion recognition of 4 basic emotions (relaxed, happy, sad, angry) on valence dimension, arousal dimension or valence-arousal synthesis
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
