Pith. sign in

REVIEW 3 major objections 6 minor 39 references

Contactless Cardiac Pulse Monitoring Using Event Cameras

T0 review · 3 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

Pith's one-line read Event-camera face recordings preserve cardiac pulse well enough for contactless heart-rate monitoring.

desk verdict First face-based event-camera pulse extraction with a CNN, and the code release is real; but the headline 120 FPS advantage over RGB rests on ten subjects and one split, so treat the comparison as preliminary. read the letter →

arxiv 2505.09529 v2 pith:AIVT7BBR submitted 2025-05-14 cs.CV cs.ETcs.LGeess.IV

classification cs.CVcs.ETcs.LGeess.IV
keywords eventcamerasneuromorphicvisionremotephotoplethysmographyheartrateestimationconvolutionalneuralnetworkstemporalshiftattentionECGsignalpreprocessingcontactlesshealthmonitoring
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper sets out to show that an event camera pointed at a person's face captures enough of the cardiac pulse to estimate heart rate without contact. It trains a supervised CNN on 2D event frames built from asynchronous brightness-change events, using a processed ECG signal as the target, and reports that event-frame models match or beat an RGB-video baseline when the event frames are generated at 60 or 120 FPS. The best event model reaches 2.13 bpm RMSE, below the RGB baseline's 2.92. If the result holds, event cameras, which use little power and have high dynamic range, become a plausible low-power sensor for remote health monitoring.

What carries the argument

The machinery is the event frame: events, each carrying timestamp, coordinates, and polarity, are binned into non-overlapping windows of fixed duration and summed by polarity into a single-channel 2D image, making the asynchronous event stream compatible with frame-based network architectures. A modified TS-CAN, a two-branch convolutional attention network with temporal shift modules, takes the same single-channel event frame into both branches and learns to output a normalized signal difference that post-processing integrates into a pulse waveform. On the label side, the ECG signal is inverted to emphasize s-wave peaks, smoothed, bandpass filtered, clipped, resampled to the frame rate, and differentiated, so the supervision target is designed to align with the blood-pulse timing the event stream is assumed to encode.

What would settle it

Simultaneously record an event camera, a synchronized RGB camera, and both ECG and fingertip SpO2 from the same subjects; train the same model with ECG-derived labels, then compare its predicted pulse peaks against SpO2 pulse peaks beat by beat. If the event-frame predictions do not track SpO2 timing within a fraction of a cardiac cycle while the RGB model does, the claim that cardiac information is preserved in event streams would be falsified.

Watch

Extended reading notes

Core claim

The paper claims that physiological cardiac information in the facial region is effectively preserved within the event stream, and that a simple 2D event-frame representation lets a standard convolutional network recover it. On its local dataset, the event-frame model at 120 FPS reaches RMSE 2.13 bpm, MAE 1.58 bpm, and Pearson correlation 0.97, outperforming the 30 FPS RGB model (RMSE 2.92) and the 60 FPS event model (RMSE 2.54). The 30 FPS event model is slightly worse than RGB but still within the accepted error range for remote photoplethysmography. The authors also report that an event-count-only method from prior work fails on face data, with RMSE above 25 bpm, while their spatial event-frame method does not.

Load-bearing premise

The paper's central claim rests on the assumption that the processed ECG signal used as the training label tracks the timing and shape of the optical blood-volume pulse that the event camera sees in the face; if that mapping is wrong, the reported errors measure something other than cardiac information in the event stream.

Editorial extensions

If this is right

  • Contactless pulse monitoring can run on event-camera hardware, which offers microsecond-level latency and lower power than frame-based video capture.
  • Existing frame-based remote-PPG network architectures can be reused for event data simply by feeding them temporally binned event frames, lowering the barrier to neuromorphic health sensing.
  • Increasing the event-frame sampling rate from 30 to 120 FPS improves heart-rate accuracy in the reported experiments, so shorter temporal bins appear to be preferable for this task.
  • The failure of event-count-only methods on face data suggests that spatial information is necessary for face-based event pulse estimation.
  • The achieved RMSE values fall within the accepted threshold of 5 bpm for remote-PPG methods, so the accuracy is in the clinically relevant range.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the finding generalizes to natural head motion, event cameras could be built into vehicle cabins or wearables without some of the illumination constraints that affect RGB remote-PPG, because event cameras have high dynamic range.
  • A direct test of the label assumption would be to compare the model's predicted pulse peaks against a simultaneous fingertip SpO2 waveform, whose timing reflects blood arrival in the extremities rather than electrical heart activity.
  • Synthetic generation of event streams from public RGB remote-PPG datasets could be used to scale training data and to test whether the model learns the same optical blood-volume signal that RGB methods exploit.
  • Beat-by-beat waveform comparison, not just heart-rate error, would clarify whether event frames preserve pulse shape or only periodicity.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. This paper investigates whether event cameras can capture cardiac pulse information from face recordings. The authors collect a proprietary dataset of 64 subjects with simultaneous event, RGB, and ECG data, represent event streams as 2D event frames, and train a modified TS-CAN to predict a processed ECG signal. They compare models trained on RGB frames and on event frames at 30, 60, and 120 FPS, evaluating heart rate via FFT of the post-processed prediction. Reported RMSEs are 2.92 bpm for RGB 30 FPS, and 3.32, 2.54, and 2.13 bpm for event frames at 30, 60, and 120 FPS, respectively. The abstract and conclusion interpret the 120 FPS event result as outperforming the RGB baseline and supporting event cameras as a viable low-power contactless cardiac monitoring sensor.

Significance. If the result holds, this is a useful first demonstration: it extends event-camera vitals monitoring from wrist and chest to the face, uses end-to-end learning rather than hand-crafted event-count features, and provides code and trained weights for reproducibility. The design choices (held-out subject evaluation, no subject exclusion, standard r-PPG error metrics) are appropriate for a feasibility study. The main limitation is statistical: the headline comparison rests on 10 test subjects, one split, and one training run per configuration, with highly variable per-subject errors.

major comments (3)
  1. [Section 4.4, Tables 3 and 4] The claim that the 120 FPS event model outperforms the RGB baseline (RMSE 2.13 vs. 2.92 bpm) is not statistically established. The evaluation uses a single random split (44/10/10 subjects), one training run per configuration, and a test set of only 10 subjects. Per-subject RMSE in Table 4 varies from 0.00 to 5.05 for the 120 FPS event model, so the aggregate difference could be dominated by one or two subjects. Please report confidence intervals, a paired significance test over subjects, and ideally multiple splits or seeds before drawing the comparative conclusion in the abstract.
  2. [Section 3.2] The label pre-processing inverts the ECG to emphasize s-wave peaks and then applies smoothing, band-pass filtering, clipping, resampling, and differentiation, but no evidence is given that the resulting signal tracks the facial blood-volume pulse observed by the event camera. Since heart rate is derived from the FFT peak of the post-processed output, waveform phase details may be less critical, but the claim that the model extracts 'physiological cardiac information' would be considerably stronger if the processed ECG were validated against an SpO2 or standard r-PPG reference, or if an ablation without the inversion step were reported.
  3. [Section 5, Conclusion] The manuscript concludes that event cameras are a viable low-power alternative to RGB for remote heart rate monitoring, but the current evidence supports a narrower feasibility claim: on a stationary, proprietary 10-subject test set, a 120 FPS event-frame model achieves competitive aggregate RMSE. Given the acknowledged lack of motion scenarios, the small test set, and the single-split evaluation, I recommend softening the conclusion to a feasibility demonstration until broader evaluation is available.
minor comments (6)
  1. [Section 3.1.1] Equation (1) defines Wj with an ambiguous set-builder notation; please clarify that i and k are the first and last event indices in the window and write the set condition explicitly.
  2. [Section 3.1.1] The pre-processing description says frames are normalized between -8 and 8 and then multiplied by 255 and cast to 8-bit unsigned integers; negative values cannot be represented in this conversion, so please describe the actual normalization and typecasting procedure.
  3. [Figure 3] The caption says 'A 10-second sample is inverted and smoothed' while the text says the figure shows a 1-second sample; these statements are inconsistent.
  4. [Section 4.3] The paper states that validation loss was used to select optimal weights, but no validation metrics or selected epoch numbers are reported; please include them.
  5. [Section 5, Table 5] The implementation details of the event-count baseline from [26] (filter order, cutoff frequencies, FFT windowing) are not provided, which limits reproducibility of the comparison.
  6. [Section 4.1] The sentence about some subjects providing an additional minute of break data is vague; clarify how recording durations varied across the 64 subjects.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the event-to-pulse mapping is learned from independent ECG labels and evaluated on held-out subjects.

full rationale

The paper's derivation chain is empirical rather than definitional: event frames are generated from raw event streams, a modified TS-CAN is trained with MSE against an ECG-derived label signal, and heart rate is then computed from the post-processed model output via FFT. The labels come from an independent ECG sensor, not from the event stream or from the model's own output, so the learned mapping is tested rather than assumed. The ECG preprocessing in Section 3.2 (inversion, smoothing, bandpass filtering, clipping, resampling, and differentiation) is a modeling assumption about label suitability, but it does not encode the event-frame content or the final heart-rate value, so it is not circular. No fitted parameter is renamed as a prediction, no uniqueness theorem is imported from the authors' prior work, and the self-citations (e.g., Refs. [2], [22], [24]) serve as literature context or external comparison baselines rather than load-bearing justification for the central claim. The limitations acknowledged in Sections 4.4 and 5 (proprietary data, stationary subjects, small test set, single split) are concerns about statistical strength and external validity, not about circularity. The honest finding is therefore a non-finding on circularity, with score 0.

Assumptions & free parameters 8 free parameters · 4 assumptions · 0 invented entities

The central claim is empirical and depends on several hand-chosen preprocessing and training settings, plus domain assumptions about ECG label validity, fixed face cropping, and event frame fidelity. No new physical entities are introduced. The largest unstated burden is the assumption that the inverted and filtered ECG label behaves like the optical pulse signal in the face.

free parameters (8)
  • Event frame window period L = 33333 us (30 FPS), 16666 us (60 FPS), 8333 us (120 FPS)
    Hand-chosen temporal bin width for event frames; it defines the frame rate and directly controls the central comparison between 30, 60, and 120 FPS event models.
  • Event downsampling factor df = 5
    Hand-chosen downsampling factor reducing 1280x720 event coordinates to 144x144 event frames; no ablation is reported.
  • RGB face crop and downsample size = 400x400 crop, then 144x144
    Fixed-index cropping and downsampling for the RGB baseline; hand-chosen without reported sensitivity analysis.
  • ECG Savitzky-Golay smoothing window = 101 samples, quadratic fit
    Hand-chosen smoothing parameters for the 1 kHz ECG signal used to generate training labels.
  • ECG bandpass filter cutoffs = 0.0015 and 0.0048 normalized (0.75 Hz and 2.5 Hz)
    Hand-chosen first-order Butterworth cutoffs based on standard heart rate frequency limits.
  • ECG clipping percentiles = Top and bottom 1% clipped
    Hand-chosen outlier clipping applied to the processed ECG label signal.
  • Event frame normalization range = [-8, 8], then multiplied by 255 and cast to uint8
    Chosen as 'best-practice for visual clarity and noise reduction'; the cast from signed floating point to unsigned integer is not fully specified.
  • Training hyperparameters = learning rate 18e-5 or 9e-5; frame depth 10, 20, or 40; batch size 8; 30 epochs
    Hand-chosen settings. Frame depth is scaled with frame rate, and learning rate is halved for the 60 and 120 FPS experiments, so the frame-rate comparison is not a clean single-variable change.
assumptions (4)
  • standard math FFT-based heart rate extraction from a bandpass-filtered signal identifies the dominant cardiac frequency.
    Used in the post-processing pipeline in Section 3.3 and in the comparison with [26]; this is a standard assumption in rPPG signal processing.
  • domain assumption The processed and inverted ECG signal is a valid proxy for the blood-volume pulse signal visible in the face.
    Section 3.2 states the ECG is inverted to emphasize s-wave peaks and filtered to resemble SpO2 signals. The paper does not independently validate that this processed ECG label matches the optical pulse signal in timing or shape.
  • domain assumption A fixed-index face crop captures the face without per-frame tracking, which is adequate for the recorded stationary subjects.
    Section 3.1.1 applies fixed cropping directly to event coordinates, and Section 4.1 describes subjects sitting still with eyes open or closed. This assumption would break under head motion.
  • domain assumption Accumulating event polarities into 2D frames preserves enough spatial-temporal structure for a CNN to extract cardiac information.
    This premise is the core modeling choice behind the event frame representation in Section 3.1.1. The experimental results support it on this dataset, but the paper does not compare event frames with other event representations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Contactless Cardiac Pulse Monitoring Using Event Cameras." pith.science (2026). https://pith.science/paper/AIVT7BBR

@misc{pith2026250509529,
  author       = {Pith},
  title        = {Pith review of: Contactless Cardiac Pulse Monitoring Using Event Cameras},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AIVT7BBR}},
  note         = {Machine review of arXiv:2505.09529}
}
read the original abstract

Time event cameras are a novel technology for recording scene information at extremely low latency and with low power consumption. Event cameras output a stream of events that encapsulate pixel-level light intensity changes within the scene, capturing information with a higher dynamic range and temporal resolution than traditional cameras. This study investigates the contact-free reconstruction of an individual's cardiac pulse signal from time event recording of their face using a supervised convolutional neural network (CNN) model. An end-to-end model is trained to extract the cardiac signal from a two-dimensional representation of the event stream, with model performance evaluated based on the accuracy of the calculated heart rate. The experimental results confirm that physiological cardiac information in the facial region is effectively preserved within the event stream, showcasing the potential of this novel sensor for remote heart rate monitoring. The model trained on event frames achieves a root mean square error (RMSE) of 3.32 beats per minute (bpm) compared to the RMSE of 2.92 bpm achieved by the baseline model trained on standard camera frames. Furthermore, models trained on event frames generated at 60 and 120 FPS outperformed the 30 FPS standard camera results, achieving an RMSE of 2.54 and 2.13 bpm, respectively.

Figures

Figures reproduced from arXiv: 2505.09529 by the authors.

Figure 1
Figure 1. Preparation of time event data for remote cardiac pulse estimation. The acquired events are divided into [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Comparison between ECG and SpO2 data acquired during the same period at a sampling rate of 1 kHz. The [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Illustration of pre-processing steps applied to the ECG signal to obtain the signal used to label frames. A [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: The R and S waves on a standard ECG peak. [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Diagram of the modified TS-CAN trained on event frames. As opposed to the original TS-CAN model where [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Sample output from the different r-PPG models covering a ten-second duration. Both the model output [PITH_FULL_IMAGE:figures/full_fig_p011_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 25 canonical work pages

  1. [26]

    Non-invasive vital parameter detection using neuromorphic cameras exploring the use of neuromor- phic cameras for biomedical applications

    Niklas Huhs. Non-invasive vital parameter detection using neuromorphic cameras exploring the use of neuromor- phic cameras for biomedical applications. 11 2024

  2. [1]

    Backpropagation applied to handwritten zip code recognition

    Yann LeCun, Bernhard Boser, John S Denker, Donnie Henderson, Richard E Howard, Wayne Hubbard, and Lawrence D Jackel. Backpropagation applied to handwritten zip code recognition. Neural computation, 1(4):541– 551, 1989

  3. [2]

    Event cameras in automotive sensing: A review

    Waseem Shariff, Mehdi Sefidgar Dilmaghani, Paul Kielty, Mohamed Moustafa, Joe Lemley, and Peter Corcoran. Event cameras in automotive sensing: A review. IEEE Access, 2024

  4. [3]

    Pushing the limits of asynchronous graph-based object detection with event cameras

    Daniel Gehrig and Davide Scaramuzza. Pushing the limits of asynchronous graph-based object detection with event cameras. arXiv preprint arXiv:2211.12324, 2022

  5. [4]

    Spiking-fer: spiking neural network for facial expression recognition with event cameras

    Sami Barchid, Benjamin Allaert, Amel Aissaoui, José Mennesson, and Chaabane C Djeraba. Spiking-fer: spiking neural network for facial expression recognition with event cameras. In Proceedings of the 20th International Conference on Content-based Multimedia Indexing, pages 1–7, 2023

  6. [5]

    E-gaze: Gaze estimation with event camera

    Nealson Li, Muya Chang, and Arijit Raychowdhury. E-gaze: Gaze estimation with event camera. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024

  7. [6]

    Event-based vision: A survey

    Guillermo Gallego, Tobi Delbrück, Garrick Orchard, Chiara Bartolozzi, Brian Taba, Andrea Censi, Stefan Leutenegger, Andrew J Davison, Jörg Conradt, Kostas Daniilidis, et al. Event-based vision: A survey. IEEE transactions on pattern analysis and machine intelligence, 44(1):154–180, 2020

  8. [7]

    Recent event camera innovations: A survey

    Bharatesh Chakravarthi, Aayush Atul Verma, Kostas Daniilidis, Cornelia Fermuller, and Yezhou Yang. Recent event camera innovations: A survey. arXiv preprint arXiv:2408.13627, 2024

Show all 39 references
  1. [8]

    Remote plethysmographic imaging using ambient light

    Wim Verkruysse, Lars O Svaasand, and J Stuart Nelson. Remote plethysmographic imaging using ambient light. Optics express, 16(26):21434–21445, 2008

  2. [9]

    Multi-task temporal shift attention networks for on-device contactless vitals measurement

    Xin Liu, Josh Fromm, Shwetak Patel, and Daniel McDuff. Multi-task temporal shift attention networks for on-device contactless vitals measurement. Advances in Neural Information Processing Systems, 33:19400–19411, 2020

  3. [10]

    Non-contact ppg signal and heart rate estimation with multi-hierarchical convolutional network

    Bin Li, Panpan Zhang, Jinye Peng, and Hong Fu. Non-contact ppg signal and heart rate estimation with multi-hierarchical convolutional network. Pattern Recognition, 139:109421, 2023

  4. [11]

    Robust pulse rate from chrominance-based rppg

    Gerard De Haan and Vincent Jeanne. Robust pulse rate from chrominance-based rppg. IEEE Transactions on Biomedical Engineering, 60(10):2878–2886, 2013

  5. [12]

    Deepphys: Video-based physiological measurement using convolutional attention networks

    Weixuan Chen and Daniel McDuff. Deepphys: Video-based physiological measurement using convolutional attention networks. In Proceedings of the european conference on computer vision (ECCV), pages 349–365, 2018

  6. [13]

    Attention is all you need

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. Advances in neural information processing systems, 30, 2017

  7. [14]

    Tsm: Temporal shift module for efficient video understanding

    Ji Lin, Chuang Gan, and Song Han. Tsm: Temporal shift module for efficient video understanding. In Proceedings of the IEEE/CVF international conference on computer vision, pages 7083–7093, 2019

  8. [15]

    Remote photoplethysmograph signal measurement from facial videos using spatio-temporal networks

    Zitong Yu, Xiaobai Li, and Guoying Zhao. Remote photoplethysmograph signal measurement from facial videos using spatio-temporal networks. arXiv preprint arXiv:1905.02419, 2019. 13 A PREPRINT - AUGUST 21, 2025

  9. [16]

    A logical calculus of the ideas immanent in nervous activity

    Warren S McCulloch and Walter Pitts. A logical calculus of the ideas immanent in nervous activity. The bulletin of mathematical biophysics, 5:115–133, 1943

  10. [17]

    Rhythmformer: Extracting rppg signals based on hierarchical temporal periodic transformer

    Bochao Zou, Zizheng Guo, Jiansheng Chen, and Huimin Ma. Rhythmformer: Extracting rppg signals based on hierarchical temporal periodic transformer. arXiv preprint arXiv:2402.12788, 2024

  11. [18]

    Instantaneous physiological estimation using video transformers

    Ambareesh Revanur, Ananyananda Dasari, Conrad S Tucker, and László A Jeni. Instantaneous physiological estimation using video transformers. In Multimodal AI in healthcare: A paradigm shift in health intelligence , pages 307–319. Springer, 2022

  12. [19]

    Physformer: Facial video-based physiological measurement with temporal difference transformer

    Zitong Yu, Yuming Shen, Jingang Shi, Hengshuang Zhao, Philip HS Torr, and Guoying Zhao. Physformer: Facial video-based physiological measurement with temporal difference transformer. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pages 41...

  13. [20]

    Physformer++: Facial video-based physiological measurement with slowfast temporal difference transformer

    Zitong Yu, Yuming Shen, Jingang Shi, Hengshuang Zhao, Yawen Cui, Jiehua Zhang, Philip Torr, and Guoy- ing Zhao. Physformer++: Facial video-based physiological measurement with slowfast temporal difference transformer. International Journal of Computer Vision, 131(6):1307–1330, 2023

  14. [21]

    Eddd: Event-based drowsiness driving detection through facial motion analysis with neuromorphic vision sensor

    Guang Chen, Lin Hong, Jinhu Dong, Peigen Liu, Jörg Conradt, and Alois Knoll. Eddd: Event-based drowsiness driving detection through facial motion analysis with neuromorphic vision sensor. IEEE Sensors Journal , 20(11):6170–6181, 2020

  15. [22]

    Real-time face & eye tracking and blink detection using event cameras

    Cian Ryan, Brian O’Sullivan, Amr Elrasad, Aisling Cahill, Joe Lemley, Paul Kielty, Christoph Posch, and Etienne Perot. Real-time face & eye tracking and blink detection using event cameras. Neural Networks, 141:87–97, 2021

  16. [23]

    Spikepoint: An efficient point-based spiking neural network for event cameras action recognition.arXiv preprint arXiv:2310.07189, 2023

    Hongwei Ren, Yue Zhou, Yulong Huang, Haotian Fu, Xiaopeng Lin, Jie Song, and Bojun Cheng. Spikepoint: An efficient point-based spiking neural network for event cameras action recognition.arXiv preprint arXiv:2310.07189, 2023

  17. [24]

    Heart rate detection using an event camera

    Aniket Jagtap, RamaKrishna Venkatesh Saripalli, Joe Lemley, Waseem Shariff, and Alan F Smeaton. Heart rate detection using an event camera. In 2023 IEEE International Symposium on Multimedia (ISM), pages 243–246. IEEE, 2023

  18. [25]

    Periodicity intensity reveals insights into time series data: three use cases

    Alan F Smeaton and Feiyan Hu. Periodicity intensity reveals insights into time series data: three use cases. Algorithms, 16(2):119, 2023

  19. [27]

    The Fourier transform and its applications, volume 31999

    Ronald Newbold Bracewell and Ronald N Bracewell. The Fourier transform and its applications, volume 31999. McGraw-Hill New York, 1986

  20. [28]

    A 128×128 120 db 15µs latency asynchronous temporal contrast vision sensor

    Patrick Lichtsteiner, Christoph Posch, and Tobi Delbruck. A 128×128 120 db 15µs latency asynchronous temporal contrast vision sensor. IEEE journal of solid-state circuits, 43(2):566–576, 2008

  21. [29]

    Heart rate estimation network from facial videos using spatiotem- poral feature image

    Kokila Bharti Jaiswal and Toshanlal Meenpal. Heart rate estimation network from facial videos using spatiotem- poral feature image. Computers in Biology and Medicine, 151:106307, 2022

  22. [30]

    Smoothing and differentiation of data by simplified least squares procedures

    Abraham Savitzky and Marcel JE Golay. Smoothing and differentiation of data by simplified least squares procedures. Analytical chemistry, 36(8):1627–1639, 1964

  23. [31]

    Deep physiological sensing toolbox

    Xin Liu, Xiaoyu Zhang, Girish Narayanswamy, Yuzhe Zhang, Yuntao Wang, Shwetak Patel, and Daniel McDuff. Deep physiological sensing toolbox. arXiv preprint arXiv:2210.00716, 2022

  24. [32]

    Plux biosignals

    PLUX Biosignals. Plux biosignals. https://www.pluxbiosignals.com/. Accessed: 2024-07-23

  25. [33]

    Fixing weight decay regularization in adam.arXiv preprint arXiv:1711.05101, 5, 2017

    Ilya Loshchilov, Frank Hutter, et al. Fixing weight decay regularization in adam.arXiv preprint arXiv:1711.05101, 5, 2017

  26. [34]

    Super-convergence: Very fast training of neural networks using large learning rates

    Leslie N Smith and Nicholay Topin. Super-convergence: Very fast training of neural networks using large learning rates. In Artificial intelligence and machine learning for multi-domain operations applications, volume 11006, pages 369–386. SPIE, 2019

  27. [35]

    Non-contact, automated cardiac pulse measurements using video imaging and blind source separation

    Ming-Zher Poh, Daniel J McDuff, and Rosalind W Picard. Non-contact, automated cardiac pulse measurements using video imaging and blind source separation. Optics express, 18(10):10762–10774, 2010

  28. [36]

    Zenodo, September 2023

    Facial Camera-Based Heart Rate Estimation Using r-PPG Convolutional Neural Networks. Zenodo, September 2023

  29. [37]

    Unsupervised skin tissue segmentation for remote photoplethysmography

    Serge Bobbia, Richard Macwan, Yannick Benezeth, Alamin Mansouri, and Julien Dubois. Unsupervised skin tissue segmentation for remote photoplethysmography. Pattern Recognition Letters, 124:82–90, 2019. 14 A PREPRINT - AUGUST 21, 2025

  30. [38]

    Non-contact video-based pulse rate measurement on a mobile service robot

    Ronny Stricker, Steffen Müller, and Horst-Michael Gross. Non-contact video-based pulse rate measurement on a mobile service robot. InThe 23rd IEEE International Symposium on Robot and Human Interactive Communication, pages 1056–1062. IEEE, 2014

  31. [39]

    Remote photoplethysmography for heart rate measurement: A review

    Hanguang Xiao, Tianqi Liu, Yisha Sun, Yulin Li, Shiyi Zhao, and Alberto Avolio. Remote photoplethysmography for heart rate measurement: A review. Biomedical Signal Processing and Control, 88:105608, 2024. 15

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.