Facial Camera-Based Heart Rate Estimation Using r-PPG Convolutional Neural Networks

This paper aims to study whether the heart signal loss is necessarily an indicator of heart rate prediction performance of supervised deep learning models trained to extract the heart signal (r-PPG signal) from face video. We present the findings of our study, where we trained and tested three state-of-the-art deep neural convolutional network architectures on publicly available datasets. It was submitted to the 25th Irish Machine Vision and Image Processing Conference (IMVIP) and was accepted for a special session on Deep Learning in Medical Image Processing.The research conducted in this publication was funded by the Irish Research Council under project ID EBPPG/2021/92 as a part of the Employment-Based Programme Postgraduate Scholarship in partnership with the Xperi Corporation (FotoNation)