AI-Assisted Remote Health Monitoring Using Remote Photoplethysmography Signals for Stress Detection and Vital Sign Alerting
DOI:
https://doi.org/10.63503/acset.128Keywords:
rPPG, remote health monitoring, vital signs, stress detection, deep learning, signal processing, ResNet, GRUAbstract
Photoplethysmography (PPG) is a basic technology that enables non-invasive vital sign monitoring through traditional contact pulse oximeters that measure blood volume and heart rate. The need for physical contact limits traditional PPG systems from operating continuously, as they require direct contact, which creates patient discomfort during long monitoring sessions and leads to skin-related problems and hygiene concerns. Remote Photoplethysmography (rPPG) is a contactless method that uses ambient light and standard RGB cameras to capture facial video data for physiological signal detection. Technology fails to reach its full potential because real-world environments introduce motion artifacts and changing lighting conditions. This makes clinical alerting data unreliable. The research introduces a new framework that links theoretical concepts to actual implementation methods. Our approach begins with multiple facial regions analyses to obtain baseline data, which we then feed into the Dilated Attention-ResNet-BiGRU deep learning network. The system learns to convert distorted visual inputs into precise waveforms that match the quality of contact sensors used as gold-standard references. The clean signal enables HRV calculation and estimation of physiological stress levels because it blocks both illumination disturbances and movement-related noise. This technology offers strong potential for ongoing surveillance, benefiting senior citizens and patients who need medical supervision after surgical procedures. The system delivers fast computational power while generating precise vital signs and providing an operational, real-time system that alerts users to stress and health abnormalities.
References
[1] A.C. Sumathi, et al., ”Health Monitoring System For Pregnant Women,” in ICEIET 2016, 2016.
[2] J. Yang, et al., ”PPG biometric recognition with singular value decom- position...,” J. Intell. Fuzzy Syst., vol. 43, pp. 3599-3610, 2022.
[3] E. Karapinar and E. Sevinc, ”A non-invasive heart rate prediction method using a convolutional approach,” Med. Biol. Eng. Comput., vol. 63, pp. 901-914, 2025.
[4] M. T. Islam, et al., ”SPECMAR: fast heart rate estimation from PPG signal...,” Med. Biol. Eng. Comput., vol. 57, pp. 689-702, 2019.
[5] S. Lee, et al., ”Ultra-short-term stress measurement using RGB camera- based remote photoplethysmography...,” Med. Biol. Eng. Comput., vol. 63, pp. 497-510, 2025. A. Qayyum, et al., ”Assessment of physiological states from contactless face video...,” Computing, vol. 105, pp. 761-781, 2023.
[6] P. Pirzada, et al., ”Remote Photoplethysmography (rPPG): A State-of- the-Art Review,” medRxiv, 2023.
[7] S. Chen, et al., ”An image enhancement based method for improving rPPG extraction...,” Biomed. Signal Process. Control, vol. 100, p. 106963, 2025.
[8] Q. Zhu, et al., ”A Comparative Study of Principled rPPG-Based Pulse Rate Tracking Algorithms for Fitness Activities,” IEEE Trans. Biomed. Eng., vol. 72, no. 1, pp. 152-164, 2025.
[9] L. Q. Abdulrahaman, ”Two-Stage Motion Artifact Reduction Algorithm for rPPG Signals...,” Arab. J. Sci. Eng., vol. 49, pp. 2925-2933, 2023.
[10] M. H. Chowdhury, et al., ”LGI-rPPG-Net: A shallow encoder-decoder model for rPPG signal estimation...,” Biomed. Signal Process. Control, vol. 89, p. 105687, 2024.
[11] Lokendra and G. Puneet, ”AND-rPPG: A novel denoising-rPPG network for improving remote heart rate estimation,” Comput. Biol. Med., vol. 141, p. 105146, 2022.
[12] M. Javidh, et al., ”CHROM-Y: Illumination-Adaptive Robust Remote Photoplethysmography...,” Signals, vol. 6, no. 4, p. 72, 2025.
[13] F. Haugg, M. Elgendi, and C. Menon, ”GRGB rPPG: An Efficient Low- Complexity Remote Photoplethysmography-Based Algorithm for Heart Rate Estimation,” Bioengineering, vol. 10, no. 2, p. 243, 2023.
[14] S. Qaadan, G. Al Jayyousi, and A. Alkhalaileh, ”A Unified Bench- marking Framework for Classical Machine Learning Based Heart Rate Estimation from RGB and NIR rPPG,” Electronics, vol. 15, no. 2, p. 218, 2026.
[15] Z. Yu, et al., ”Remote Heart Rate Measurement from Highly Compressed Facial Videos: an End-to-end Deep Learning Solution with Video Enhancement,” in ICCV, 2019.
[16] K. Egorov, et al., ”Gaze into the Heart: A Multi-View Video Dataset for rPPG and Health Biomarkers Estimation,” in MM ’25, 2025.
[17] F. Haugg, M. Elgendi, and C. Menon, ”Effectiveness of Remote PPG Construction Methods: A Preliminary Analysis,” Bioengineering, vol. 9, no. 10, p. 485, 2022.
[18] S. Guler, et al., ”Optimal digital filter selection for remote photoplethys- mography (rPPG) signal conditioning,” Biomed. Phys. Eng. Express, vol. 9, p. 027001, 2023.
[19] P. Toye, ”VitalVideos-Worldwide: A large and diverse rPPG dataset with rich ground truths,” in Proc. IEEE/CVF Int. Conf. Comput. Vis. (ICCV) Workshops, 2023.
Downloads
Published
Conference Proceedings Volume
Section
License
Copyright (c) 2026 Adroid Conference Series: Engineering and Technology
This work is licensed under a Creative Commons Attribution 4.0 International License.
