Deep Learning-Based Image Enhancement for Robust Remote Photoplethysmography in Various Illumination Scenarios

Shutao Chen, Sui Kei Ho, Jing Wei Chin, Kin Ho Luo, Tsz Tai Chan, Richard H.Y. So, Kwan Long Wong; Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2023, pp. 6077-6085

Abstract


Remote photoplethysmography (rPPG) is a non-invasive and convenient approach for measuring human vital signs using a camera. However, accurate measurement can be challenging due to the different illumination of the surrounding environment. In this study, we present a deep learning-based image enhancement model (IEM) inspired by the Retinex theory to improve the robustness of rPPG signal extraction and heart rate (HR) estimation in different lighting conditions. We finetuned the IEM with a time-shifted negative Pearson correlation between the PPG signal ground truth and the predicted rPPG signal from a pretrained 3D CNN (PhysNet). We evaluated our method using a publicly available dataset (BH-rPPG) of various lighting scenarios and our own private dataset. Our results demonstrate that our proposed model is generalizable and significantly improves rPPG extraction and HR estimation accuracy across a range of illumination intensities.

Related Material


[pdf]
[bibtex]
@InProceedings{Chen_2023_CVPR, author = {Chen, Shutao and Ho, Sui Kei and Chin, Jing Wei and Luo, Kin Ho and Chan, Tsz Tai and So, Richard H.Y. and Wong, Kwan Long}, title = {Deep Learning-Based Image Enhancement for Robust Remote Photoplethysmography in Various Illumination Scenarios}, booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops}, month = {June}, year = {2023}, pages = {6077-6085} }