Heart Rate Estimation Research Articles

Hybrid Neural Network Models to Estimate Vital Signs from Facial Videos

Introduction: Remote health monitoring plays a crucial role in telehealth services and the effective management of patients, which can be enhanced by vital sign prediction from facial videos. Facial videos are easily captured through various imaging devices like phone cameras, webcams, or surveillance systems. Methods: This study introduces a hybrid deep learning model aimed at estimating heart rate (HR), blood oxygen saturation level (SpO2), and blood pressure (BP) from facial videos. The hybrid model integrates convolutional neural network (CNN), convolutional long short-term memory (convLSTM), and video vision transformer (ViViT) architectures to ensure comprehensive analysis. Given the temporal variability of HR and BP, emphasis is placed on temporal resolution during feature extraction. The CNN processes video frames one by one while convLSTM and ViViT handle sequences of frames. These high-resolution temporal features are fused to predict HR, BP, and SpO2, capturing their dynamic variations effectively. Results: The dataset encompasses 891 subjects of diverse races and ages, and preprocessing includes facial detection and data normalization. Experimental results demonstrate high accuracies in predicting HR, SpO2, and BP using the proposed hybrid models. Discussion: Facial images can be easily captured using smartphones, which offers an economical and convenient solution for vital sign monitoring, particularly beneficial for elderly individuals or during outbreaks of contagious diseases like COVID-19. The proposed models were only validated on one dataset. However, the dataset (size, representation, diversity, balance, and processing) plays an important role in any data-driven models including ours. Conclusions: Through experiments, we observed the hybrid model’s efficacy in predicting vital signs such as HR, SpO2, SBP, and DBP, along with demographic variables like sex and age. There is potential for extending the hybrid model to estimate additional vital signs such as body temperature and respiration rate.

FMCW-based contactless heart rate monitoring

Heart disease is a significant global health issue. Traditional methods for heart rate monitoring typically require close physical contact, which limits the continuity and convenience of monitoring. To achieve real-time, non-contact heartbeat monitoring, researchers have introduced millimeter-wave radar technology. The technology’s penetration and privacy offer a potential solution for heart condition monitoring. Therefore, this study utilized frequency-modulated continuous wave (FMCW) radar for heart rate monitoring. Firstly, the collected millimeter-wave radar signals were preprocessed to accurately locate the area of cardiac activity in the human body. Secondly, an adaptive variational mode decomposition (A-VMD) algorithm was designed to extract the heartbeat signal, considering signal variations caused by random body movements and respiration and their harmonics, to obtain an accurate heartbeat signal. Finally, the accurate heart rate is obtained by weighted estimation based on the harmonic relationship of the heartbeat signal. The study invited ten subjects to participate in the experiment to verify the effectiveness of this method. The results show that this method can reduce the influence of and random body movements and respiration and harmonics on heart rate monitoring, the average absolute error of heart rate estimation is less than four bpm.

Low-Complexity Timing Correction Methods for Heart Rate Estimation Using Remote Photoplethysmography.

With the rise of modern healthcare monitoring, heart rate (HR) estimation using remote photoplethysmography (rPPG) has gained attention for its non-contact, continuous tracking capabilities. However, most HR estimation methods rely on stable, fixed sampling intervals, while practical image capture often involves irregular frame rates and missing data, leading to inaccuracies in HR measurements. This study addresses these issues by introducing low-complexity timing correction methods, including linear, cubic, and filter interpolation, to improve HR estimation from rPPG signals under conditions of irregular sampling and data loss. Through a comparative analysis, this study offers insights into efficient timing correction techniques for enhancing HR estimation from rPPG, particularly suitable for edge-computing applications where low computational complexity is essential. Cubic interpolation can provide robust performance in reconstructing signals but requires higher computational resources, while linear and filter interpolation offer more efficient solutions. The proposed low-complexity timing correction methods improve the reliability of rPPG-based HR estimation, making it a more robust solution for real-world healthcare applications.

From video to vital signs: a new method for contactless multichannel seismocardiography

Seismocardiography (SCG) is a technique that non-invasively measures the chest wall’s local vibrations caused by the heart’s mechanical activity. Traditionally, SCG signals have been recorded using accelerometers placed at a single location on the chest wall. This study presents an innovative, cost-effective SCG method that utilizes standard smartphone videos to capture data from multiple chest locations. The analysis of vibrations from multiple points can offer a more thorough understanding of the heart’s mechanical activity compared to signals obtained solely from a single chest location. Our approach employs computer vision and deep learning techniques to extract and improve the resolution of multichannel SCG maps obtained by video capture of chest movement. We attached a grid of patterned stickers to the chest surface and recorded videos of chest movements during different respiratory phases. Using a deep learning-based object detector and a template tracking method, we tracked the stickers across video frames and extracted the corresponding SCG signals from sticker displacements. We also developed a robust algorithm to estimate heart rate (HR) from these chest videos and identify the optimal chest location for HR estimation. The method was tested on 28 chest videos captured from 14 healthy participants. The results demonstrated that our method effectively extracted multichannel SCG maps and enhanced their resolution with a mean squared error of 0.1078 and 0.0418 for right-to-left and head-to-foot SCG signals, respectively. We observed intersubject chest vibration patterns corresponding to cardiac events including opening and closure of the heart valves. Moreover, our algorithm accurately estimated HR from 1968 SCG signals extracted from the videos compared to the gold-standard HR measured from each subject’s electrocardiogram (bias ± 1.96 SD = 0.04 ± 2.14 bpm; r = 0.99, p < 0.001). The findings from this study underscore the potential of our approach in developing a cardiac monitoring tool using a smartphone that would be widely accessible to the general public and might provide more timely detection of diseases.

Smart furniture using radar technology for cardiac health monitoring

The integration of radar technology into smart furniture represents a practical approach to health monitoring, circumventing the concerns regarding user convenience and privacy often encountered by conventional smart home systems. Radar technology’s inherent non-contact methodology, privacy-preserving features, adaptability to diverse environmental conditions, and high precision characteristics collectively establish it a compelling alternative for comprehensive health monitoring within domestic environments. In this paper, we introduce a millimeter (mm)-wave radar system positioned strategically behind a seat, featuring an algorithm capable of identifying unique cardiac waveform patterns for healthy subjects. These patterns are characterized by two peaks followed by a valley in each cycle, which can be correlated to Electrocardiogram (ECG), enabling effective cardiac waveform monitoring. The provided algorithm excels in discerning variations in heart patterns, particularly in individuals with prolonged corrected QT intervals, by minimizing high frequency breathing interference and ensuring accurate pattern recognition. Additionally, this paper addresses the influence of body movements in seated individuals, conducting a comprehensive study on heart rate variability and estimation. Experiment results demonstrate a maximum interbeat intervals (IBI) error of 30 milliseconds and an average relative error of 4.8% in heart rate estimation, showcasing the efficacy of the proposed method utilizing variational mode decomposition and a multi-bin approach.

Adaptive Estimation Algorithm for Photoplethysmographic Heart Rate Based on Finite State Machine

In order to address the issue of heart rate susceptibility to motion artifacts (MAs) when extracting it from photoplethysmography (PPG) signals, a heart rate estimation algorithm based on the finite state machine (FSM) is proposed. The algorithm first applies band-pass filtering to the PPG and three-axis acceleration signals. The strength of MA is assessed based on the acceleration data. If a strong MA is detected, recursive least squares (RLS) filtering is applied; otherwise, it is omitted. Then, the signal is subjected to an empirical wavelet transform (EWT). Based on the EWT results, the current state is identified, and the corresponding spectral peak screening method is selected to estimate the heart rate. The mean absolute errors of the algorithm on 12 sets of public data and 8 sets of testing data are 0.93 and 1.76 beats per minute (bpm), respectively. The results of the experiment show that compared with other dominant algorithms, the proposed algorithm estimates heart rate with a smaller mean absolute error and can extract heart rate more effectively.

MFF-Net: A Lightweight Multi-Frequency Network for Measuring Heart Rhythm from Facial Videos.

Remote photo-plethysmography (rPPG) is a useful camera-based health motioning method that can measure the heart rhythm from facial videos. Many well-established deep learning models can provide highly accurate and robust results in measuring heart rate (HR) and heart rate variability (HRV). However, these methods are unable to effectively eliminate illumination variation and motion artifact disturbances, and their substantial computational resource requirements significantly limit their applicability in real-world scenarios. Hence, we propose a lightweight multi-frequency network named MFF-Net to measure heart rhythm via facial videos in a short time. Firstly, we propose a multi-frequency mode signal fusion (MFF) mechanism, which can separate the characteristics of different modes of the original rPPG signals and send them to a processor with independent parameters, helping the network recover blood volume pulse (BVP) signals accurately under a complex noise environment. In addition, in order to help the network extract the characteristics of different modal signals effectively, we designed a temporal multiscale convolution module (TMSC-module) and spectrum self-attention module (SSA-module). The TMSC-module can expand the receptive field of the signal-refining network, obtain more abundant multiscale information, and transmit it to the signal reconstruction network. The SSA-module can help a signal reconstruction network locate the obvious inferior parts in the reconstruction process so as to make better decisions when merging multi-dimensional signals. Finally, in order to solve the over-fitting phenomenon that easily occurs in the network, we propose an over-fitting sampling training scheme to further improve the fitting ability of the network. Comprehensive experiments were conducted on three benchmark datasets, and we estimated HR and HRV based on the BVP signals derived by MFF-Net. Compared with state-of-the-art methods, our approach achieves better performance both on HR and HRV estimation with lower computational burden. We can conclude that the proposed MFF-Net has the opportunity to be applied in many real-world scenarios.

TransformHR: A Novel Approach to Athlete Heart Rate Monitoring With Transformer‐Based Wearable Technology

ABSTRACTThis study presents an innovative Transformer‐based deep learning model for heart rate estimation in wearable devices, addressing accuracy challenges during high‐intensity activities. Our model, utilizing self‐attention mechanisms and energy‐efficient computational strategies, achieves a mean absolute error (MAE) of 4.0–4.2 beats per minute and correlation coefficients of 0.92–0.93 on public datasets PPG‐DaLiA and WESAD. On our more challenging proprietary athlete dataset, the model maintains excellent performance with an MAE of 4.3 beats per minute and a correlation coefficient of 0.91. Ablation studies demonstrate the importance of attention mechanisms, positional encoding, and Gaussian‐weighted loss functions to model performance. These results highlight the potential of our approach in improving heart rate monitoring accuracy in wearable technologies, particularly for high‐intensity athletic data, showcasing its adaptability and effectiveness in real‐world athletic scenarios.

Sensor-Based Frailty Assessment Using Fitbit.

This study evaluated the reliability of Fitbit in assessing frailty based on motor and heart rate (HR) parameters through a validated upper extremity function (UEF) test, which involves 20 s of rapid elbow flexion. For motor performance, participants completed six trials of full elbow flexion using their right arm, with and without weight. Fitbit and a commercial motion sensor were worn on the right arm. For HR measurements, an ECG system was placed on the left chest alongside the Fitbit on the left wrist. Motor parameters assessing speed, flexibility, weakness, exhaustion, and HR before, during, and after UEF were measured. A total of 42 participants (age = 22 ± 3) were recruited. For motor parameters, excellent agreement was observed between the wearable sensor and Fitbit, except for flexibility (ICC = 0.87 ± 0.09). For HR parameters, ICC values showed weak agreement between ECG and Fitbit for HR increase and recovery (ICC = 0.24 ± 0.11), while moderate to stronger agreement was seen for mean HR during baseline, task, and post-task (ICC = 0.81 ± 0.13). Fitbit is a reliable tool for assessing frailty through motor parameters and provides reasonably accurate HR estimates during baseline, task, and recovery periods. However, Fitbit's ability to track rapid HR changes during activity is limited.

Reliability Of Maximum Heart Rate Estimation Equations In Athletes In The Field Football Base Category

The objective of this study was to verify the reliability of the HRmax prediction equations proposed by Fox, Tanaka and Nikolaidis in field football athletes who belong to the youth categories of a club in the first division of Brazilian league. One hundred and sixty-four players, aged between 14 and 20, took part in this study. In order to determine the athletes' HRmax (bpm), Yo-Yo Endurance Test Level 1 was used and defined as the highest value achieved during the test. Data were expressed as mean, standard deviation (SD) of the mean, minimum and maximum values.Independent Student's t-test and Pearson's correlation coefficient were used in statistical analyses. The results showed that none of the three available equations (Fox, Tanaka and Nikolaidis) provide accurate HRmax values in a sample of youth football players, underestimating or overestimating HRmax values. The equation that provided the most accurate HRmax was the Nikolaidis equation. This result is probably due to the fact that this equation was developed based on an example of a football player.

A cascaded approach of group sparse mode decomposition and deep neural network for heart rate estimation using reference signal-less PPG signal

A cascaded approach of group sparse mode decomposition and deep neural network for heart rate estimation using reference signal-less PPG signal

Contactless radar-based heart rate estimation in palliative care – a feasibility study and possible use in symptom management

BackgroundHeart rate (HR) monitoring is a medical standard to provide information about a patient’s health status. In palliative care, relationship and social engagement are crucial therapeutic concepts. For fear of disrupting communication, social contact, and care, continuous HR monitoring is underutilised despite its potential to inform on symptom burden and therapeutic effects. This study investigates radar-based HR monitoring as an innovative and burden-free approach for palliative care patients, compares its accuracy with conventional ECG methods, and shows potential for therapeutic guidance.MethodsA single-centre, comparative clinical trial was conducted with palliative care patients at the ward of the Department of Palliative Medicine of the University Hospital of Erlangen. The HR measurements obtained with radar were compared with Holter ECG (study arm I, overnight) and Task Force® Monitor (TFM)-based ECG validation recordings (study arm II, one hour). In addition, long-term radar measurements without validation were analysed in comparison with clinical health records (study arm III).ResultsBoth validation methods showed correlation by scatter plot, modified Bland-Altman plot, and equivalence testing. N = 34 patients participated in study arm I. HR of 4,079 five-minute intervals was analysed. Radar measurements and ECG showed high agreement: difference of HRs was within \\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\:\\pm\\:$$\\end{document}5 bpm in 3780 of 4079 (92.67%) and within ±13.4 bpm (\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\:\\pm\\:$$\\end{document}1.96 times the SD of the mean) in 3979 (97.55%) intervals, respectively. In study arm II, n = 19 patients participated. 57,048 heart beats were analysed. The HR difference was within \\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\:\\pm\\:$$\\end{document}5 bpm for 53,583 out of 57,048 beats (93.93%) and within \\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\:\\pm\\:$$\\end{document}8.2 bpm ( ± 1.96 times the SD of the mean) in 55,439 beats (97.25%), respectively. Arm III showed HR changes extracted from radar data in correlation with symptoms and treatment.ConclusionRadar-based HR monitoring shows a high agreement in comparison with ECG-based HR monitoring and thus offers an option for continuous and above all burden-free HR assessment, with the potential for use in symptom management in palliative care, among others. Further research and technological advancements are still necessary to fully realize this innovative approach in enhancing palliative care practices.

Integrating Remote Photoplethysmography and Machine Learning on Multimodal Dataset for Noninvasive Heart Rate Monitoring.

Non-contact heart monitoring is crucial in advancing telemedicine, fitness tracking, and mass screening. Remote photoplethysmography (rPPG) is a non-contact technique to obtain information about heart pulse by analyzing the changes in the light intensity reflected or absorbed by the skin during the blood circulation cycle. However, this technique is sensitive to environmental lightning and different skin pigmentation, resulting in unreliable results. This research presents a multimodal approach to non-contact heart rate estimation by combining facial video and physical attributes, including age, gender, weight, height, and body mass index (BMI). For this purpose, we collected local datasets from 60 individuals containing a 1 min facial video and physical attributes such as age, gender, weight, and height, and we derived the BMI variable from the weight and height. We compare the performance of two machine learning models, support vector regression (SVR) and random forest regression on the multimodal dataset. The experimental results demonstrate that incorporating a multimodal approach enhances model performance, with the random forest model achieving superior results, yielding a mean absolute error (MAE) of 3.057 bpm, a root mean squared error (RMSE) of 10.532 bpm, and a mean absolute percentage error (MAPE) of 4.2% that outperforms the state-of-the-art rPPG methods. These findings highlight the potential for interpretable, non-contact, real-time heart rate measurement systems to contribute effectively to applications in telemedicine and mass screening.

Multi-source partial domain adaptation with Gaussian-based dual-level weighting for PPG-based heart rate estimation

Photoplethysmography (PPG) signals from wearable devices have expanded the accessibility of heart rate estimation. Recent advances in deep learning have significantly improved the generalizability of heart rate estimation from PPG signals. However, these models exhibit performance degradation when used for new subjects with different PPG distributions. Although previous studies have attempted subject-specific training and fine-tuning techniques, they require labeled data for each new subject, limiting their practicality. In response, we explore the application of domain adaptation techniques using only unlabeled PPG signals from the target subject. However, naive domain adaptation approaches do not adequately account for the variability in PPG signals among different subjects in the training dataset. Furthermore, they overlook the possibility that the heart rate range of the target subject may only partially overlap with that of the source subjects. To address these limitations, we propose a novel multi-source partial domain adaptation method, GAussian-based dUaL-level weighting (GAUL), designed for the PPG-based heart rate estimation, formulated as a regression task. GAUL considers and adjusts the contribution of relevant source data at the domain and sample levels during domain adaptation. The experimental results on three benchmark datasets demonstrate that our method outperforms existing domain adaptation approaches, enhancing the heart rate estimation accuracy for new subjects without requiring additional labeled data. The code is available at: https://github.com/Im-JihyunKim/GAUL.

Accelerometer-based heart rate adjustment for ambulatory stress research.

Using heart rate (HR) measurements to detect mental stress in naturalistic settings is hampered by the physiological impact of hemodynamic and metabolic demands. Correcting HR for these demands can help isolate fluctuations in HR associated with psychosocial stress responses, a concept termed additional heart rate (aHR). This study examined whether adding predictors for posture, activity type, and lagged movement intensity for the prolonged impact of physical activity (PA) improved aHR estimation across various manipulations of mental stress, posture, and PA in a controlled laboratory environment (n = 197). Accelerometer signals were used to obtain the movement intensity and to classify posture and activity type. Posture, activity type, and lagged movement intensity each led to a significant improvement in HR estimation, as measured by adjusted R2 and root mean squared error. However, HR was overestimated during quiet sitting. The aHR, computed as the difference between observed and predicted HR, generally underestimated observed task-baseline reactivity but was sensitive to individual differences in reactivity to mental stressors. Between-subject correlations of aHR with task-baseline reactivity ranged from 0.62 to 0.93 across conditions. On a within-subject level, the ability of aHR to differentiate between exposure to physical stress and mental stress was limited (recall = 0.32, precision = 0.31), but better than that of observed HR (recall = 0.02, precision = 0.02). Future research should explore the potential of this novel aHR estimation method in differentiating physical and mental demands on HR in daily life, and its predictive value for health outcomes.

A non-invasive heart rate prediction method using a convolutional approach.

The research focuses on leveraging convolutional neural networks (CNNs) to enhance the analysis of physiological signals, specifically photoplethysmogram (PPG) data which is a valuable tool for non-invasive heart rate prediction. Recognizing the global challenge of heart failure, the study seeks to provide a rapid, accurate, and non-invasive alternative to traditional, uncomfortable blood pressure cuffs. To achieve more accurate and efficient heart rate estimates, a k-fold CNN model with an optimal number of convolutional layers is employed. While the findings show promising results, the study addresses potential sources of error in cuffless PPG-based heart rate measurement, including motion artifacts and skin color variations, emphasizing the need for validation against gold standard methods. The research optimizes a CNN model with optimal layers, operating on 1D arrays of 8-s data slices and employing k-fold cross-validation to mitigate probabilistic uncertainties. Finally, the model yields a remarkable minimum absolute error (MAE) rate of 6.98 beats per minute (bpm), marking a significant 10% improvement over recent studies. The study also advances medical diagnostics and data analysis, then lays a strong foundation for developing cost-effective, reliable devices that offer a more comfortable and efficient way of predicting heart rate.

Clinical Validation of a Custom Wearable Patch for Accurate and Comfortable Vital Sign Monitoring in Pediatric Patients

Background/Purpose: Measuring vital signs in pediatric patients requires special consideration and adaptation due to varying anatomy and wide age range. In addition, children's anxiety, uncooperativeness, and high activity levels further complicate measurements, necessitating devices and algorithms designed to minimize the inaccuracies and discomfort. In this work, the performance of a custom wearable patch mounted on the mid-sternum was validated in uncontrolled settings on a cohort including 84 pediatric patients. Methods: Three-minute-long electrocardiogram (ECG), seismocardiogram (SCG) and photoplethysmogram (PPG) signals were acquired using the custom patch. First, pre-processing and signal smoothing algorithms were employed to suppress the out-of-band and motion noise. Two different tasks were then studied: (i) Heart rate (HR) and respiration rate were derived from the ECG, PPG and SCG signals individually. During HR derivation from the SCG, a novel Teager-energy-based HR estimation algorithm was proposed. (ii) Clinical relevance of the SCG signals was shown through mapping the SCG characteristics to body mass index (BMI) and blood pressure values. Results: While the best HR estimation was achieved through the PPG-infrared signal with an absolute error of 2.22.1 bpm, the best respiration estimation was achieved with PPG-Red signal with an error of 2.62.2 breaths/min. On the other hand, regression models resulted in a minimum of 85% confidence interval, revealing that the SCG characteristics indeed have salient correlation with the BMI and blood pressure values. Conclusion: Overall, such a system can potentially be leveraged in clinical practices to achieve more comfortable and accurate measurements.

Non-contact monitoring of human cardiorespiratory activity during sleep using FMCW millimeter wave radar

Millimeter wave radar can be utilized for non-contact, long-term, and accurate health monitoring. This paper presents an application for vital signs detection including respiratory rate (RR) and heart rate (HR) during sleep, as well as facilitating rapid detection and extended recording of apnea syndrome. This paper employs frequency spectrum analysis for RR estimation and develops an algorithm of template matching estimation (TME) for real-time, accurate HR estimation by comparing radar signal fragments to a template with a frequency signature. The time required for HR estimation is reduced to 1 s. In addition, a neural network-based model was developed for detecting the apnea events with the accuracy of 93.60% and calculates the duration in seconds. The system underwent testing using irregular analog signals of vital signs. In the actual scene testing, the accuracy of the RR estimation is 90.58%, and 90.25% for HR estimation.

FEHAT: efficient, large scale and automated heartbeat detection in Medaka fish embryos.

High-resolution imaging of model organisms allows the quantification of important physiological measurements. In the case of fish with transparent embryos, these videos can visualize key physiological processes, such as heartbeat. High throughput systems can provide enough measurements for the robust investigation of developmental processes as well as the impact of system perturbations on physiological state. However, few analytical schemes have been designed to handle thousands of high-resolution videos without the need for some level of human intervention. We developed a software package, named FEHAT, to provide a fully automated solution for the analytics of large numbers of heart rate imaging datasets obtained from developing Medaka fish embryos in 96-well plate format imaged on an Acquifer machine. FEHAT uses image segmentation to define regions of the embryo showing changes in pixel intensity over time, followed by the classification of the most likely position of the heart and Fourier Transformations to estimate the heart rate. Here, we describe some important features of the FEHAT software, showcasing its performance across a large set of medaka fish embryos and compare its performance to established, less automated solutions. FEHAT provides reliable heart rate estimates across a range of temperature-based perturbations and can be applied to tens of thousands of embryos without the need for any human intervention. Data used in this manuscript will be made available on request.

Heart Rate Estimation Algorithm Integrating Long and Short-Term Temporal Features

Non-contact heart rate monitoring from facial videos utilizing remote photoplethysmography (rPPG) has gained significant traction in remote health monitoring. Given that rPPG captures the dynamic blood flow within the human body and constitutes a time-series signal characterized by periodic properties, this study introduced a three-dimensional convolutional neural network (3D CNN) designed to simultaneously address long-term periodic and short-term temporal characteristics for effective rPPG signal extraction. Firstly, differential operations are employed to preprocess video data, enhancing the face’s dynamic features. Secondly, building upon the 3D CNN framework, multi-scale dilated convolutions and self-attention mechanisms were integrated to enhance the model’s temporal modeling capabilities further. Finally, interpolation techniques are applied to refine the heart rate calculation methodology. The experiments conducted on the UBFC-rPPG dataset indicate that, compared with the existing optimal algorithm, the average absolute error (MAE) and the root mean square error (RMSE) realized significant enhancements of approximately 28% and 35%. Additionally, through comprehensive analyses such as cross-dataset experiments and complexity analyses, the validity and stability of the proposed algorithm in the task of heart rate estimation were manifested.