Malasinghe, L. PKatsigiannis, SRamzan, N2022-03-112022-03-112018-05-16978-1-4503-6366-2https://rda.sliit.lk/handle/123456789/1576Remote and contactless heart rate detection is still an open research issue of great clinical importance. Available approaches lack the necessary accuracy and reliability for acceptance by medical experts. In this study, we propose a new method for remote heart rate extraction using the Microsoft KinectTM v2.0 image sensor. The proposed approach relies on signal processing and machine learning methods in order to create a model for accurate estimation of the heart rate via RGB and infrared face videos. Electrocardiography (ECG) recordings and RGB and infrared face videos, captured using the KinectTM v2.0 image sensor, were acquired from 17 subjects and used to create a machine learning model for remote heart rate detection. Experimental evaluation through supervised regression experiments showed that the proposed approach achieved a mean absolute error of 6.972 bpm, demonstrating the capabilities of the underlying technology.enHeart rateRemote heart rate extractionRemote patient monitoringKinect v2.0RGB sensorInfrared sensorFast fourier transformArticle10.1145/3232059.3232060