FOREX Press I. J. of Electrical & Electronics Research
Support Open Access
  • Home/
  • IJEER-130410

Research Article |

Hybrid CNN-BiLSTM Autoencoder for Anomaly Detection in Remote Photoplethysmography(rPPG) Signals

Author(s): Moussa Mmadi1*, George Kamucha2, and Ciira wa Maina3

Published In : International Journal of Electrical and Electronics Research (IJEER)        Volume 13, Issue 4

Publisher : FOREX Publication

Published : 10 December 2025

e-ISSN : 2347-470X

Page(s) : 704-711

DOI: https://doi.org/10.37391/IJEER.130410

Abstract

Remote photoplethysmography (rPPG) is becoming increasingly popular as a non-contact method for tracking physiological parameters like heart rate and respiration rate. However, the accuracy of rPPG signals is often compromised by various factors, including movement, lighting variations, and sensor noise. These challenges can severely impact signal quality, leading to unreliable measurements and hindering the practical application of rPPG-based systems. In this research, we introduced and assessed the effectiveness of a hybrid Convolutional Neural Network (CNN) and Bidirectional Long Short-Term Memory (BiLSTM) Autoencoder model specifically designed for reliable anomaly detection in rPPG time-series data. This model aims to detect anomalies such as sensor noise and motion artifacts within the remote photoplethysmography signal. The UBFC-rPPG, COHFACE, and the PURE datasets were utilized for training and testing, demonstrating excellent performance in distinguishing clean segments from noisy ones, with high precision, recall, F1-score, and low false positive rates.

Keywords: Remote Photoplethysmography, Convolutional Neural Network, Anomaly Detection, Time-Series, BiLSTM.




Moussa Mmadi*, Department of Electrical Engineering, Pan African University Institute for Basic Sciences, Technology and Innovation, Nairobi, Kenya; Email: moussa.mmadi@students.jkuat.ac.ke

George Kamucha, Department of Electrical and Information Engineering, University of Nairobi, Nairobi, Kenya; Email: gkamucha@uonbi.ac.ke

Ciira wa Maina, Department of Electrical Engineering, Dedan Kimathi University of Technology, Nyeri, Kenya; Email: ciira.maina@dkut.ac.ke

    [1] Guangkun Nie, et al. (2024). A review of deep learning methods for photoplethysmography data. arXiv preprint arXiv:2401.12783.
    [2] Shutao Chen, Kwan-Long Wong, Jing-Wei Chin, Tsz-Tai Chan, and Richard H. Y. So (2024). DiffPhys: Enhancing Signal-to-Noise Ratio in Remote Photoplethysmography Signal Using a Diffusion Model Approach. Bioengineering, 11(8), 743. DOI: 10.3390/bioengineering11080743.
    [3] Lee, K., Kim, S., An, B., Seo, H., Park, S., and Lee, E. C. (2023). Noise-Assessment-Based Screening Method for Remote Photoplethysmography Estimation. Applied Sciences, 13(17), 9818.
    [4] J. Speth, A. D. Bagheri, D. McDuff, and S. Hershey, "Hallucinated Heartbeats: Anomaly-Aware Remote Pulse Estimation," arXiv preprint arXiv:2303.06452, 2023.
    [5] Botina-Monsalve, Deivid, Yannick Benezeth, and Johel Miteran. "Performance analysis of remote photoplethysmography deep filtering using long short-term memory neural network." BioMedical Engineering Online 21.1 (2022): 69.
    [6] Varun Chandola, Arindam Banerjee, and Vipin Kumar (2009). Anomaly Detection: A Survey. ACM Computing Surveys (CSUR), 41(3), 1–58.
    [7] Massaroni, C., Nicolo, A., Sacchetti, M., and Schena, E. (2020). Contactless Methods for Measuring Respiratory Rate: A Review. IEEE Sensors Journal, 21(11), 12821–12839.
    [8] Lee, J., Kim, M., Park, H. K., and Kim, I. Y. (2020). Motion Artifact Reduction in Wearable Photoplethysmography Based on Multi-Channel Sensors with Multiple Wavelengths. Sensors, 20(5), 1493. DOI: 10.3390/s20051493.
    [9] S.-Q. Liu and P. C. Yuen, "Robust Remote Photoplethysmography Estimation with Environmental Noise Disentanglement," IEEE Transactions on Image Processing, vol. 32, pp. 2570–2583,2023.
    [10] Choi, K., Yi, J., Park, C., and Yoon, S. (2021), Deep Learning for Anomaly Detection in Time-Series Data: Review, Analysis, and Guidelines. IEEE Access, 9, 120043–120065. DOI: 10.1109/ACCESS.2021.3107975.
    [11] Fahim, M., and Sillitti, A. (2019). Anomaly Detection, Analysis and Prediction Techniques in IoT Environment: A Systematic Literature Review. IEEE Access, 7, 81664–81681. DOI: 10.1109/ACCESS.2019.2924225.
    [12] R. Yousefi, M. Nourani, Separating arterial and venous-related components of photoplethysmographic signals for accurate extraction of oxygen saturation and respiratory rate, IEEE J. Biomed. Health Inf. 19 (3) (2014) 848–857.
    [13] A. Gudi, M. Bittner, J. van Gemert, Real-time webcam heart-rate and variability estimation with clean ground truth for evaluation, Appl. Sci. 10 (23) (2020) 8630.
    [14] G. de Haan and V. Jeanne, "Robust Pulse Rate from Chrominance-Based rPPG," IEEE Transactions on Biomedical Engineering, vol. 60, no. 10, pp. 2878–2886, 2013.
    [15] M.-Z. Poh, D. J. McDuff, and R. W. Picard, "Non-contact, automated cardiac pulse measurements using video imaging and blind source separation," Optics Express, vol. 18, no. 10, pp. 10762–10774, 2010.
    [16] S. Guler, A. Golparvar, O. Ozturk, H. Dogan, and M. K. Yapici, “Optimal digital filter selection for remote photoplethysmography (rPPG) signal conditioning,” Biomedical Physics & Engineering Express, vol. 9, no. 2, p. 025003, Feb. 2023.
    [17] Stricker, R., Müller, S., Gross, H.-M. “Non-contact Video-based Pulse Rate Measurement on a Mobile Service Robot,” in: Proc. 23rd IEEE Int. Symposium on Robot and Human Interactive Communication (Ro-Man 2014), Edinburgh, Scotland, UK, pp. 1056-1062, IEEE 2014.
    [18] Wang, Y., Perry, M., Whitlock, D., and Sutherland, J. W. (2022). Detecting Anomalies in Time Series Data from a Manufacturing System Using Recurrent Neural Networks. Journal of Manufacturing Systems, 62, 823–834. DOI: 10.1016/j.jmsy.2022.01.011.
    [19] R. Chalapathy and S. Chawla, “Deep Learning for Anomaly Detection: A Survey,” Jan. 2019, Accessed: Feb. 23, 2023. [Online]. Available: http://arxiv.org/abs/1901.03407.
    [20] W. Verkruysse, L. O. Svaasand, and J. S. Nelson, "Remote plethysmographic imaging using ambient light," Optics Express, vol. 16, no. 26, pp. 21434–21445, 2008.
    [21] A. Hyvärinen and E. Oja, "Independent component analysis: algorithms and applications," Neural Networks, vol. 13, no. 4-5, pp. 411–430, 2000.
    [22] S. Bobbia, R. Macwan, Y. Benezeth, A. Mansouri, J. Dubois, "Unsupervised skin tissue segmentation for remote photoplethysmography", Pattern Recognition Letters, 2017.
    [23] Pankaj, Malhotra. "LSTM-based encoder-decoder for multi-sensor anomaly detection." CoRR 1607 (2016).
    [24] M. Mmadi, G. N. Kamucha, C. Maina, “Contactless Algorithm for Identifying the Facial Area Containing the Most Vital Sign-Related Information Using Machine Learning,” Harbin Engineering Journal, vol.46, no. 05, pp. 89-99, May. 2024, https://harbinengineeringjournal.com/index.php/journal/article/view/4200/2375.
    [25] Keras Docs: Anomaly Detection in Time Series Using Autoencoders, Accessed: May. 10, 2025. [Online]. Available: https://keras.io/examples/timeseries/timeseries_anomaly_detection/
    [26] Nawaz, Ali, Shehroz S. Khan, and Amir Ahmad. "Ensemble of autoencoders for anomaly detection in biomedical data: A narrative review." IEEE Access 12 (2024): 17273-17289.
    [27] Guillaume Heusch, André Anjos, Sébastien Marcel, “A reproducible study on remote heart rate measurement”, arXiv, 2016. [online]. Available: http://publications.idiap.ch/index.php/publications/show/3688
    [28] "PURE dataset," [Online]. Available: http://www.tu-ilmenau.de/neurob/data-sets/pulse. [Accessed: Apr. 30, 2024].

Moussa Mmadi, George Kamucha, and Ciira wa Maina (2025), Hybrid CNN-BiLSTM Autoencoder for Anomaly Detection in Remote Photoplethysmography(rPPG) Signals. IJEER 13(4), 704-711. DOI: 10.37391/IJEER.130410.