Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

Pulse-PPG: An Open-Source Field-Trained PPG Foundation Model for Wearable Applications Across Lab and Field Settings

T0 review · 3 major / 5 minor · reviewed 2026-08-09 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper argues that a PPG foundation model trained only on raw, uncurated field data from 120 smartwatch users outperforms a clinical-data-trained model on 10 of 11 health tasks.

desk verdict A solid, well-documented open-source PPG encoder with a fair benchmark against PaPaGei, but the headline claim that field data itself beats clinical data is confounded by corpus scale and needs a matched-size retest. read the letter →

arxiv 2502.01108 v2 pith:M7QGIHAP submitted 2025-02-03 cs.LG cs.AIeess.SP

Mithun Saha , Maxwell A. Xu , Wanting Mao , Sameer Neupane , James M. Rehg , Santosh Kumar This is my paper · ORCID
classification cs.LGcs.AIeess.SP
keywords photoplethysmographyfoundationmodelwearablehealthcontrastivelearningself-supervisedfieldPPGrelativemotif-baseddistance
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 establish that a PPG (photoplethysmography) foundation model can be built from raw, uncurated smartwatch data collected in daily life, rather than from clean clinical recordings, and that this field-trained model transfers better to both wearable and clinical tasks. It introduces Pulse-PPG, an open-source encoder pre-trained on about 200 million seconds of wrist PPG from 120 participants over 100 days. Across 11 downstream tasks on five datasets, Pulse-PPG outperforms PaPaGei, a state-of-the-art open-source PPG foundation model trained on clinical data, on 10 of 11 tasks; the only exception is sleep disturbance, a domain absent from the field pre-training data. The paper also shows that pre-training the same architecture on a large clinical PPG dataset is worse on 10 of 11 tasks, suggesting that realistic field variability, not signal cleanliness, is what makes the representations transfer.

What carries the argument

The load-bearing object is the learnable motif-based distance function used to compare PPG windows without segmenting beats. A motif is a short temporal shape within the pulsative waveform, such as a systolic rise; the distance function uses a cross-attention reconstruction error in which each motif of the anchor window retrieves the closest motif in the candidate window through a softmax kernel regression and tries to reconstruct itself. Because it is trained on masked field PPG, it learns to match motifs even in noisy, unsegmented signals. The frozen distance function then drives the relative contrastive loss: for every anchor, candidates are ranked by distance, the closest becomes the positive pair, and farther candidates become negative pairs, so the encoder learns fine-grained relative similarity instead of coarse binary similarity. The encoder is a 1D ResNet-26 with instance normalization and global pooling that maps variable-length PPG into a 512-dimensional embedding, and its 127k-parameter distance model keeps the pre-training tractable.

What would settle it

A decisive check would be to take clean clinical PPG beats, compute the learned motif distances between all pairs of windows, and see whether the closest-ranked pairs share independent beat-morphology labels (for example, systolic rise time or dicrotic notch position) significantly more often than random pairs. If the rankings carry no such physiological signal, the pre-training objective is measuring artifacts, not pulse shape, and the reported transfer gains would not be expected to generalize.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that noise is information. Pulse-PPG is trained in two self-supervised stages: first, a lightweight dilated-convolution network learns an unsupervised motif-based distance function by reconstructing a masked two-second span of a PPG window from the most similar motifs in another window; second, a 28.5-million-parameter 1D ResNet encoder is trained with a relative contrastive loss that uses this frozen distance to order a candidate set of same-subject and cross-subject windows by their distance from an anchor, pulling the embedding toward relatively close windows and away from relatively distant ones. Frozen embeddings from this encoder, evaluated with linear probes, beat PaPaGei's clinical-data embeddings on 10 of 11 tasks spanning wearable field stress and activity, wearable lab stress and instantaneous heart rate, and clinical blood pressure, hypertension, and sleep disturbance. Re-training the same model on a curated MIMIC-III clinical PPG corpus produces worse results on 10 of 11 tasks, including several clinical tasks. The paper reads this as evidence that exposure to real-world motion artifacts, ambient light, and skin-contact variability teaches a PPG encoder fine-grained, transferable structure that clean clinical data does not provide.

Load-bearing premise

The load-bearing assumption is that the unsupervised motif-based distance function orders PPG windows by real physiological similarity rather than by noise patterns or reconstruction artifacts; if that ranking is semantically empty, the relative contrastive loss trains the encoder to sort noise, and the downstream gains would not survive contact with new datasets.

Editorial extensions

If this is right

  • An open-source, field-trained PPG encoder gives researchers a general-purpose backbone; tasks with small labeled datasets can be solved with a linear probe on frozen embeddings instead of training from scratch.
  • Pre-training data selection matters more than data cleanliness: field PPG beats clinical PPG even for clinical downstream tasks, so future PPG foundation models should be trained on realistic wearable recordings rather than only curated hospital waveforms.
  • The one clear failure, sleep disturbance, coincides with a domain absent from pre-training, implying that coverage of target physiology matters at least as much as the pre-training objective.
  • Because Pulse-PPG has 28.5M parameters and still ranks in the top two against Chronos (200M) and MOMENT (385M) on average metrics, a modest PPG-specific model can compete with much larger general time-series foundation models on physiological tasks.
  • Fine-tuning Pulse-PPG improves average F1 by 9.6% and average MAE by 19.78% over linear probing, providing a cheap path to task-specific performance.

Reading between the lines

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

  • Beyond the paper, the same relative-contrastive recipe could transfer to other biosignals (ECG, EMG, respiration) that lack geometric invariances, as long as a domain-specific motif distance can be learned from masked reconstruction.
  • Beyond the paper, the results imply a testable ranking rule for pre-training data: match deployment noise conditions first, match deployment labels second; a model trained on wrist PPG from one device generation should generalize better to a new wrist device than one trained on finger clinical PPG.
  • Beyond the paper, the motif-distance function could be validated directly against physiological ground truth by checking whether windows ranked 'closest' share beat-morphology features (systolic rise time, dicrotic notch position) on clean clinical data; the paper does not perform that check.
  • Beyond the paper, adding overnight PPG from a sleep study to the field pre-training corpus would be a natural extension, and the paper's own field-versus-clinical result predicts that the mixed corpus would beat clinical-only pre-training even on sleep tasks.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper introduces Pulse-PPG, a 28.5M-parameter PPG foundation model pre-trained with a relative contrastive learning objective (RelCon) on uncurated 4-minute windows of wrist-worn field PPG from the MOODS study (120 participants, up to 100 days). The authors evaluate frozen embeddings with linear probing and fine-tuning across 11 tasks in five datasets spanning wearable field, wearable lab, and clinical PPG. They report that Pulse-PPG outperforms the open-source clinical PPG foundation model PaPaGei on 10/11 tasks, that field pre-training beats clinical pre-training on the same architecture, and that the released model is a useful general-purpose backbone. The paper also includes ablations on window length, normalization, and a parameter-matched 'Light Pulse-PPG' model.

Significance. If the claims hold, this is a useful contribution to PPG foundation modeling: it provides an open-source model trained on realistic field data, a relatively clean linear-probing benchmark against PaPaGei, and a set of downstream evaluations spanning multiple domains. The authors are also transparent about limitations (label noise in field stress, lack of skin-tone data, single-study pretraining). The central scientific claim—that field-data pretraining itself, rather than scale or architecture, drives the gains—is not yet established because Experiment 3 confounds data domain with corpus size. The paper ships code and weights, which is a concrete strength that should be credited.

major comments (3)
  1. [§6.3, Table 7] The conclusion that pre-training on field PPG outperforms pre-training on clinical PPG is confounded by corpus scale. The field model is trained on 606,833 unique 4-minute segments per epoch for 6 epochs (Section 4.3.4), while the clinical model is trained on 151,738 5-minute segments with 'the same training procedures and hyperparameters' (Section 6.3.2), i.e., the same number of epochs. The field model therefore sees roughly 4x more segments and 4x more gradient updates. The Light Pulse-PPG comparison in Appendix A.2 controls model parameters and input window length, but not pretraining data scale, so it does not resolve this confound. I ask the authors to either match the number of segments or gradient steps across the two pretraining conditions (e.g., subsample MOODS to the MIMIC-III size or extend the clinical training schedule), or to provide an analysis that explicitly separates domain from scale. As written, the paper's headline claim that 'pre-training on field data outperforms its pre-training on clinical data' is not supported by Experiment 3.
  2. [§4.1.1, Eq. (2), Eq. (5)] The pretraining pipeline assumes that the unsupervised masked-reconstruction distance function produces semantically meaningful relative orderings of PPG windows, but this assumption is never validated independently of the downstream tasks. Because this distance function defines all positive and negative relationships in RelCon (Eq. 5), a distance function that primarily encodes reconstruction artifacts or sensor noise would still be consistent with the reported downstream gains if those artifacts correlate with task labels (e.g., motion artifacts correlating with activity). I request an explicit validation or ablation: for example, compare RelCon pretraining using the learned distance against (a) random relative orderings, (b) a fixed hand-crafted PPG similarity such as normalized correlation or beat-level morphology distance, and (c) on a small labeled benchmark, an oracle label-based distance. This would test whether the motif-based ordering, rather than the contrastive framework or the field-data scale, drives the improvements.
  3. [§6.1–§6.4, Tables 6–7] No statistical significance tests, confidence intervals, or seed variance are reported for any of the headline comparisons. The '10/11 tasks' claims are based on a single run per model, and the per-task differences in Tables 6 and 7 are often small relative to the metric variability across tasks. I ask the authors to report results across at least a few random seeds (or a paired bootstrap across tasks) and to state which differences are statistically reliable. This is especially important for the field-versus-clinical comparison in Table 7, where the number of tasks is small and the effect sizes are modest for some clinical tasks.
minor comments (5)
  1. [§4.2.1 vs. §4.3.4] There is a numerical inconsistency: Section 4.2.1 says the MOODS dataset is composed of 822,247 unique 4-minute 50 Hz PPG segments from 122 participants, while Section 4.3.4 says each epoch is composed of 606,833 unique 4-minute PPG segments and that 120 participants were used. Please clarify whether the 606,833 figure excludes the validation/test splits or reflects a different exclusion criterion, and reconcile the participant counts.
  2. [Eq. (2), §3.2.2] The notation in Equation 2 is garbled: 'where X∈R^{T×D} and x∈R^D with T as the time length, ∈S as a∈ but with a subsampling of stride s' is unreadable. Please define the set S, the subsampling operator with stride s, and the dimensions of the query/key/value features precisely.
  3. [Throughout] There are several typos and formatting issues: 'comminmunity' in Section 1, 'Feasability' in the Section 8.5 heading, and the table header 'Quality Type Field Clean Clean Clean Clean' in Table 1, which is malformed.
  4. [Figures 3 and 4] The axes of Figures 3 and 4 are not labeled. Please label the horizontal and vertical axes with the metric names (e.g., F1 score, MAPE) and add units where applicable so the plots are self-contained.
  5. [§6.3.2 footnote] The clinical pretraining corpus is not identical to PaPaGei's original pretraining corpus; the paper notes that PaPaGei did not release its curation code. This limitation should be stated more prominently in the main text of Section 6.3, since it is another potential confound in the cross-model comparison.

Circularity Check

0 steps flagged · score 2.0 of 10

No circular derivation chain; the field-data advantage claim is empirically evaluated against external benchmarks, though self-cited components and validation-based model selection create a mild circularity burden.

full rationale

The paper's derivation chain is not circular by construction. The pretraining pipeline (Sections 3.2 and 4.1) trains a motif-based distance function by masked reconstruction on unlabeled field PPG (Equation 2), then uses that fixed distance to define positive and negative sets for the Relative Contrastive Loss (Equation 5), and finally trains a ResNet encoder. Downstream evaluation (Sections 6.1 through 6.4) uses linear probes and fine-tuning on held-out MOODS participants and on external datasets such as PPG-DaLiA, WESAD, SDB, and PPG-BP, so the reported predictions are not defined in terms of the pretraining objective. The RelCon and REBAR frameworks are the authors' own prior work (references 106 and 108), and the MOODS dataset (reference 70) comes from the same group; these are self-citations, but the paper supplies equations, open code, and external comparisons, so the citations are provenance rather than a load-bearing proof. The mild score of 2 reflects two burdens that are not full circularity: first, hyperparameters and early stopping were selected partly by downstream task performance (Sections 4.3.1 and 4.3.4), which can inflate apparent generalization; second, Experiment 3's field-versus-clinical pretraining comparison changes corpus scale along with domain (606,833 versus 151,738 segments per epoch), so the causal claim that field data itself drives improvement is confounded. These are validity threats, not reductions of the result to its inputs by definition.

Assumptions & free parameters 7 free parameters · 7 assumptions · 0 invented entities

The central claims rely on a chain of domain assumptions about PPG semantics, noise informativeness, label quality, and transferability of a learned distance function. No new physical entities are introduced; the motif and distance function are model constructs rather than postulated entities. The main risk is that the semantic validity of the learned distance function is assumed rather than independently verified.

free parameters (7)
  • Pre-training window length = 4 minutes
    Chosen to match 3.5 to 4.2 minute stress cycles; the ablation shows 4-minute windows perform best, but this is a hand-selected design choice that affects all downstream representations.
  • Distance-function subsampling stride s = 10
    Reduces cross-attention complexity from O(T^2) to O((T/s)^2) and affects which motifs are aligned in Equation 2.
  • Missingness mask length for distance training = 2 seconds
    Defines the reconstruction task in Section 4.1.1; no ablation is reported for this value.
  • Within-subject same-hour candidate count = 1
    Reduced from 20 in prior RelCon work to adapt to PPG; directly shapes the positive pairs used in the RelCon loss.
  • Person-specific z-normalization = applied
    Global z-score per person; the ablation shows it helps, but it is a preprocessing choice affecting embedding distributions.
  • Encoder and distance hyperparameters = kernel 15 and 11, embedding dim 64, filters 128, blocks 12, learning rates 0.001 and 0.0001, epochs 20 and 6
    Selected by parameter search on a 10-day subset as described in Section 4.3.1; these values are not derived from theory.
  • RelCon and NT-Xent temperature tau = not reported
    The temperature in Equations 4 and 5 controls contrastive sharpness, but the paper does not specify its value or tuning procedure.
assumptions (7)
  • domain assumption PPG is quasiperiodic and composed of repeating cardiac-cycle motifs, and motif differences capture semantic physiological information.
    Invoked in Section 3.2.1 as the basis for the motif-based distance function and relative contrastive learning.
  • domain assumption Noise patterns in field PPG are meaningful contextual cues correlated with activities, states, or conditions, and should be preserved rather than filtered.
    Stated in Sections 3.2.2 and 4.2.3; this assumption justifies training on uncurated data and treating noise as informative.
  • ad hoc to paper The distance function trained by masked reconstruction transfers to measuring semantic distance between distinct PPG instances across domains and durations.
    Section 4.1.1 trains the distance function on self-reconstruction and then uses it as a static function for RelCon; no independent validation of this transfer is provided.
  • domain assumption 4-minute windows capture the cyclical nature of stress-related physiological responses.
    Section 4.2.4 relies on prior stress-cycle literature to justify the pretraining input length.
  • domain assumption Within-subject, same-hour candidate sequences form semantically similar positive pairs for contrastive learning.
    Section 4.1.2 and Section 3.2.3 use this sampling assumption to define positives in the RelCon loss.
  • domain assumption Self-reported stress ratings derived with a commercial model, and activity labels generated by another model, are adequate ground truth for downstream evaluation.
    Section 8.1 and Appendix A.1 describe these label sources; noisy labels affect the MOODS downstream tasks.
  • domain assumption Representations trained on MOODS, a single US field study without race or ethnicity data, generalize to other populations and sensor types.
    Section 8.2 and Section 8.4 acknowledge the limitation; the downstream evaluation implicitly assumes this transfer.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Pulse-PPG: An Open-Source Field-Trained PPG Foundation Model for Wearable Applications Across Lab and Field Settings." pith.science (2026). https://pith.science/paper/M7QGIHAP

@misc{pith2026250201108,
  author       = {Pith},
  title        = {Pith review of: Pulse-PPG: An Open-Source Field-Trained PPG Foundation Model for Wearable Applications Across Lab and Field Settings},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/M7QGIHAP}},
  note         = {Machine review of arXiv:2502.01108}
}
read the original abstract

Photoplethysmography (PPG)-based foundation models are gaining traction due to the widespread use of PPG in biosignal monitoring and their potential to generalize across diverse health applications. In this paper, we introduce Pulse-PPG, the first open-source PPG foundation model trained exclusively on raw PPG data collected over a 100-day field study with 120 participants. Existing PPG foundation models are either open-source but trained on clinical data or closed-source, limiting their applicability in real-world settings. We evaluate Pulse-PPG across multiple datasets and downstream tasks, comparing its performance against a state-of-the-art foundation model trained on clinical data. Our results demonstrate that Pulse-PPG, trained on uncurated field data, exhibits superior generalization across clinical and mobile health applications in both lab and field settings. This suggests that exposure to real-world variability enables the model to learn fine-grained representations, making it more adaptable across tasks. Furthermore, pre-training on field data surprisingly outperforms its pre-training on clinical data in many tasks, reinforcing the importance of training on real-world, diverse datasets. To encourage further advancements in robust foundation models leveraging field data, we plan to release Pulse-PPG, providing researchers with a powerful resource for developing more generalizable PPG-based models.

Figures

Figures reproduced from arXiv: 2502.01108 by the authors.

Figure 1
Figure 1. Overview of our open-sourced∗ Pulse-PPG Foundation Model. Our model is trained on wearable field PPG with relative contrastive learning, based on the relative distances captured in our learned motif-based distance function. This model demonstrates strong performance on a wide variety of tasks across wearable field, wearable lab, and clinical settings. Photoplethysmography (PPG) in smartwatches has emerged as a widel… view at source ↗
Figure 2
Figure 2. Random 5s Snippets of PPG signals from Clinical, Wearable Lab, and Wearable Field Settings. Signal quality declines gradually but manageably when moving from clinical-grade PPG in hospital settings to wearable PPG in controlled lab conditions. However, a marked deterioration occurs in the uncontrolled real-world environment of the wearable field setting, driven by daily wear factors, such as motion artifacts [77], a… view at source ↗
Figure 3
Figure 3. Comparison of Pulse-PPG vs. PaPaGei [74], a prior PPG foundation model. The two plots show the relative performance of a linear probe evaluation for each model, with the left for classification via F1 score and the right for regression via Mean Average Percentage Error. 10/11 of the task data points reside above the slope, demonstrating how Pulse-PPG consistently outperforms PaPaGei, with particularly substantial im… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Comparison of Pre-training on Wearable Field data vs. Clinical Data to Assess Field-to-Lab Generalizability. The two plots show the relative performance of a linear probe evaluation for each model, with the left for classification via F1 score and the right for regress…
Figure 5
Figure 5. Figure 5 [PITH_FULL_IMAGE:figures/full_fig_p031_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. LSM-2: Learning from Incomplete Wearable Sensor Data

    cs.LG 2025-06 conditional novelty 6.0 of 10

    LSM-2 with Adaptive and Inherited Masking learns usable representations directly from incomplete day-long wearable data, outperforming imputation-based baselines on most tasks.

Reference graph

Works this paper leans on

121 extracted references · 54 canonical work pages · cited by 1 Pith paper

  1. [1]

    How Inarix is using DINOv2 to revolutionize the agricultural supply chain

    Accessed April, 2025. How Inarix is using DINOv2 to revolutionize the agricultural supply chain . https://ai.meta.com/blog/inarix- agricultural-supply-chain-meta-dino-v2/

    2025

  2. [2]

    https://machinelearning.apple.com/research/ introducing-apple-foundation-models

    Accessed April, 2025.Introducing Apple’s On-Device and Server Foundation Models. https://machinelearning.apple.com/research/ introducing-apple-foundation-models

    2025

  3. [3]

    Salar Abbaspourazad, Oussama Elachqar, Andrew C Miller, Saba Emrani, Udhyakumar Nallasamy, and Ian Shapiro. 2023. Large-scale training of foundation models for wearable biosignals. arXiv preprint arXiv:2312.05409 (2023)

    arXiv 2023

  4. [4]

    Tahmid Abtahi, Colin Shea, Amey Kulkarni, and Tinoosh Mohsenin. 2018. Accelerating convolutional neural network with FFT on embedded hardware. IEEE Transactions on Very Large Scale Integration (VLSI) Systems 26, 9 (2018), 1737–1749

    2018

  5. [5]

    Solaiman Ahmed, Tanveer Ahmed Bhuiyan, and Manabu Nii. 2022. PPG signal morphology-based method for distinguishing stress and non-stress conditions. Journal of Advanced Computational Intelligence and Intelligent Informatics 26, 1 (2022), 58–66

    2022

  6. [6]

    Ajmal, Tananant Boonya-Ananta, Andres J Rodriguez, VN Du Le, and Jessica C Ramella-Roman. 2021. Monte Carlo analysis of optical heart rate sensors in commercial wearables: the effect of skin tone and obesity on the photoplethysmography (PPG) signal. Biomedical optics express 12, 12 (2021), 7445–7457

    2021

  7. [7]

    Haider Ali, Imran Khan Niazi, David White, Malik Naveed Akhter, and Samaneh Madanian. 2024. Comparison of Machine Learning Models for Predicting Interstitial Glucose Using Smart Watch and Food Log. Electronics 13, 16 (2024), 3192. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol., Vol. 9, No. 3, Article 126. Publication date: September 2025. Foundatio...

    2024

  8. [8]

    John Allen. 2007. Photoplethysmography and its application in clinical physiological measurement. Physiological measurement 28, 3 (2007), R1

    2007

Show all 121 references
  1. [9]

    Moudy Sharaf Alshareef, Badraddin Alturki, and Mona Jaber. 2022. A transformer-based model for effective and exportable IoMT-based stress detection. In GLOBECOM 2022-2022 IEEE Global Communications Conference . IEEE, 1158–1163

    2022

  2. [10]

    Shun-ichi Amari. 1993. Backpropagation and stochastic gradient descent method. Neurocomputing 5, 4-5 (1993), 185–196

    1993

  3. [11]

    Maddix, Hao Wang, Michael W

    Abdul Fatir Ansari, Lorenzo Stella, Caner Turkmen, Xiyuan Zhang, Pedro Mercado, Huibin Shen, Oleksandr Shchur, Syama Sundar Rangapuram, Sebastian Pineda Arango, Shubham Kapoor, Jasper Zschiegner, Danielle C. Maddix, Hao Wang, Michael W. Mahoney, Kari Torkkola, Andrew Gordon Wi...

    2024 arXiv

  4. [12]

    Seyed Amir Hossein Aqajari, Ziyu Wang, Ali Tazarv, Sina Labbaf, Salar Jafarlou, Brenda Nguyen, Nikil Dutt, Marco Levorato, and Amir M Rahmani. 2024. Enhancing performance and user engagement in everyday stress monitoring: A context-aware active reinforcement learning approach....

    2024 arXiv

  5. [13]

    Soumyendu Banerjee and Girish Kumar Singh. 2023. A new real-time lossless data compression algorithm for ECG and PPG signals. Biomedical Signal Processing and Control 79 (2023), 104127

    2023

  6. [14]

    Rummana Bari, Md Mahbubur Rahman, Nazir Saleheen, Megan Battles Parsons, Eugene H Buder, and Santosh Kumar. 2020. Automated detection of stressful conversations using wearable physiological and inertial sensors. Proceedings of the ACM on interactive, mobile, wearable and ubiqu...

    2020

  7. [15]

    Gabriel Bénédict, Vincent Koops, Daan Odijk, and Maarten de Rijke. 2021. SigmoidF1: A smooth F1 score surrogate loss for multilabel classification. arXiv preprint arXiv:2108.10566 (2021)

    2021 arXiv

  8. [16]

    biosignalsplux. 2019. https://bio-medical.com/media/support/biosignalsplux_explorer_user_manual_v.1.0.pdf

    2019

  9. [17]

    Rishi Bommasani, Drew A Hudson, Ehsan Adeli, Russ Altman, Simran Arora, Sydney von Arx, Michael S Bernstein, Jeannette Bohg, Antoine Bosselut, Emma Brunskill, et al. 2021. On the opportunities and risks of foundation models. arXiv preprint arXiv:2108.07258 (2021)

    2021 arXiv

  10. [18]

    Xiaoli Chen, Rui Wang, Phyllis Zee, Pamela L Lutsey, Sogol Javaheri, Carmela Alcántara, Chandra L Jackson, Michelle A Williams, and Susan Redline. 2015. Racial/ethnic differences in sleep disturbances: the Multi-Ethnic Study of Atherosclerosis (MESA). Sleep 38, 6 (2015), 877–888

    2015

  11. [19]

    Yanming Chen, Chao Li, Luqi Gong, Xiang Wen, Yiwen Zhang, and Weisong Shi. 2020. A deep neural network compression algorithm based on knowledge transfer for edge devices. Computer Communications 163 (2020), 186–194

    2020

  12. [20]

    Jiho Choi, Jun Seong Lee, Moonwook Ryu, Gyutae Hwang, Gyeongyeon Hwang, and Sang Jun Lee. 2022. Attention-lrcn: long-term recurrent convolutional network for stress detection from photoplethysmography. In 2022 IEEE International Symposium on Medical Measurements and Applicatio...

    2022

  13. [21]

    Percy Cubas and Sixto Prado. 2023. Design of a PPG Signal Acquisition Platform Robust to Ambient Light. In Brazilian Technology Symposium. Springer, 206–216

    2023

  14. [22]

    Abhimanyu Das, Weihao Kong, Rajat Sen, and Yichen Zhou. 2024. A decoder-only foundation model for time-series forecasting. In Forty-first International Conference on Machine Learning

    2024

  15. [23]

    Harry J Davies, James Monsen, and Danilo P Mandic. 2024. Interpretable Pre-Trained Transformers for Heart Time-Series Data. arXiv preprint arXiv:2407.20775 (2024)

    2024 arXiv

  16. [24]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2018. Bert: Pre-training of deep bidirectional transformers for language understanding. arXiv preprint arXiv:1810.04805 (2018)

    2018 arXiv

  17. [25]

    Cheng Ding, Zhicheng Guo, Zhaoliang Chen, Randall J Lee, Cynthia Rudin, and Xiao Hu. 2024. SiamQuality: a ConvNet-based foundation model for photoplethysmography signals. Physiological Measurement 45, 8 (2024), 085004

    2024

  18. [26]

    Emadeldeen Eldele, Mohamed Ragab, Zhenghua Chen, Min Wu, Chee-Keong Kwoh, Xiaoli Li, and Cuntai Guan. 2023. Self-supervised contrastive representation learning for semi-supervised time-series classification. IEEE Transactions on Pattern Analysis and Machine Intelligence (2023)

    2023

  19. [27]

    Mohamed Elgendi. 2012. On the analysis of fingertip photoplethysmogram signals. Current cardiology reviews 8, 1 (2012), 14–25

    2012

  20. [28]

    Mohamed Elgendi, Richard Fletcher, Yongbo Liang, Newton Howard, Nigel H Lovell, Derek Abbott, Kenneth Lim, and Rabab Ward

  21. [29]

    Mohamed Elgendi, Valeria Galli, Chakaveh Ahmadizadeh, and Carlo Menon. 2022. Dataset of psychological scales and physiological signals collected for anxiety assessment using a portable device. Data 7, 9 (2022), 132

    2022

  22. [30]

    Emre Ertin, Nathan Stohs, Santosh Kumar, Andrew Raij, Mustafa Al’Absi, and Siddharth Shah. 2011. AutoSense: unobtrusively wearable sensor suite for inferring the onset, causality, and consequences of stress in the field. In Proceedings of the 9th ACM Conference on Embedded Net...

    2011

  23. [31]

    Christoph Fischer, Benno Dömer, Thomas Wibmer, and Thomas Penzel. 2016. An algorithm for real-time pulse waveform segmentation and artifact detection in photoplethysmograms. IEEE journal of biomedical and health informatics 21, 2 (2016), 372–381. Proc. ACM Interact. Mob. Weara...

    2016

  24. [32]

    Tira Nur Fitria. 2023. Artificial intelligence (AI) technology in OpenAI ChatGPT application: A review of ChatGPT in writing English essay. In ELT Forum: Journal of English Language Teaching , Vol. 12. 44–58

    2023

  25. [33]

    Giancarlo Fortino and Valerio Giampà. 2010. PPG-based methods for non invasive and continuous blood pressure measurement: An overview and development issues in body sensor networks. In 2010 IEEE International Workshop on Medical Measurements and Applications. IEEE, 10–13

    2010

  26. [34]

    Ainara Garde, Parastoo Dehkordi, Walter Karlen, David Wensley, J Mark Ansermino, and Guy A Dumont. 2014. Development of a screening tool for sleep disordered breathing in children using the phone Oximeter™. PloS one 9, 11 (2014), e112959

    2014

  27. [35]

    Mononito Goswami, Konrad Szafer, Arjun Choudhry, Yifu Cai, Shuo Li, and Artur Dubrawski. 2024. Moment: A family of open time-series foundation models. arXiv preprint arXiv:2402.03885 (2024)

    2024 arXiv

  28. [36]

    Serj Haddad, Assim Boukhayma, and Antonino Caizzone. 2020. Beat-to-beat detection accuracy using the ultra low power senbiosys PPG sensor. In European Medical and Biological Engineering Conference . Springer, 178–188

    2020

  29. [37]

    Harish Haresamudram, Irfan Essa, and Thomas Plötz. 2022. Assessing the State of Self-Supervised Human Activity Recognition Using Wearables. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol. 6, 3, Article 116 (sep 2022), 47 pages. https://doi.org/10.1145/ 3550299

    2022

  30. [38]

    Yasin Hasanpoor, Bahram Tarvirdizadeh, Khalil Alipour, and Mohammad Ghamari. 2022. Stress Assessment with Convolutional Neural Network Using PPG Signals. In 2022 10th RSI International Conference on Robotics and Mechatronics (ICRoM) . IEEE, 472–477

    2022

  31. [39]

    Jiayu He, Jianlin Ou, An He, Lin Shu, Tao Liu, Ruowen Qu, Xiangmin Xu, Zhuoming Chen, and Yifeng Yan. 2022. A new approach for daily life Blood-Pressure estimation using smart watch. Biomedical Signal Processing and Control 75 (2022), 103616

    2022

  32. [40]

    Peng He, Shaoming Meng, Yaping Cui, Dapeng Wu, and Ruyan Wang. 2023. Compression and encryption of heterogeneous signals for internet of medical things. IEEE Journal of Biomedical and Health Informatics (2023)

    2023

  33. [41]

    David T Hoffmann, Nadine Behrmann, Juergen Gall, Thomas Brox, and Mehdi Noroozi. 2022. Ranking info noise contrastive estimation: Boosting contrastive learning via ranked positives. In Proceedings of the AAAI Conference on Artificial Intelligence , Vol. 36. 897–905

    2022

  34. [42]

    Danfeng Hong, Bing Zhang, Xuyang Li, Yuxuan Li, Chenyu Li, Jing Yao, Naoto Yokoya, Hao Li, Pedram Ghamisi, Xiuping Jia, et al

  35. [43]

    Karen Hovsepian, Mustafa Al’Absi, Emre Ertin, Thomas Kamarck, Motohiro Nakajima, and Santosh Kumar. 2015. cStress: towards a gold standard for continuous stress assessment in the mobile environment. Proceedings of the 2015 ACM International Joint Conference on Pervasive and Ub...

    2015

  36. [44]

    Hangxing Hu, Jin Li, and Xiang Chen. 2022. The effect of skin melanin concentration on wrist reflectance photoplethysmography based on Monte Carlo simulation. In 2022 15th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI)...

    2022

  37. [45]

    Hyewon Jeong, Nassim Oufattole, Aparna Balagopalan, Matthew Mcdermott, Payal Chandak, Marzyeh Ghassemi, and Collin Stultz

  38. [46]

    Good Views

    Hyewon Jeong, Suyeol Yun, and Hammaad Adam. 2024. Finding" Good Views" of Electrocardiogram Signals for Inferring Abnormalities in Cardiac Condition. arXiv preprint arXiv:2411.17702 (2024)

    2024 arXiv

  39. [47]

    Shichao Kan, Yigang Cen, Yang Li, Vladimir Mladenovic, and Zhihai He. 2021. Relative order analysis and optimization for unsupervised deep metric learning. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 13999–14008

    2021

  40. [48]

    Walter Karlen, Guy Dumont, Chris Petersen, Jennifer Gow, Joanne Lim, Jules Sleiman, and J Mark Ansermino. 2011. HUMAN- CENTERED PHONE OXIMETER INTERFACE DESIGN FOR THE OPERATING ROOM-Pulse Oximeter Interfaced to a Mobile Device for Anesthesia Monitoring in the Developing World...

    2011

  41. [49]

    Ahmet Reşit Kavsaoğlu, Kemal Polat, and Mehmet Recep Bozkurt. 2016. An innovative peak detection algorithm for photoplethys- mography signals: an adaptive segmentation method. Turkish Journal of Electrical Engineering and Computer Sciences 24, 3 (2016), 1782–1796

    2016

  42. [50]

    Sungyeon Kim, Minkyo Seo, Ivan Laptev, Minsu Cho, and Suha Kwak. 2019. Deep metric learning beyond binary supervision. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition . 2288–2297

    2019

  43. [51]

    Alexander Kirillov, Eric Mintun, Nikhila Ravi, Hanzi Mao, Chloe Rolland, Laura Gustafson, Tete Xiao, Spencer Whitehead, Alexander C Berg, Wan-Yen Lo, et al. 2023. Segment anything. In Proceedings of the IEEE/CVF International Conference on Computer Vision. 4015–4026

    2023

  44. [52]

    Spyridon Kontaxis, Eduardo Gil, Vaidotas Marozas, Jesus Lazaro, Esther Garcia, Mar Posadas-de Miguel, Sara Siddi, Maria Luisa Bernal, Jordi Aguilo, Josep Maria Haro, et al. 2020. Photoplethysmographic waveform analysis for autonomic reactivity assessment in depression. IEEE Tr...

    2020

  45. [53]

    Uday Kulkarni, SM Meena, Sunil V Gurlahosur, Pratiksha Benagi, Atul Kashyap, Ayub Ansari, and Vinay Karnam. 2021. AI model compression for edge devices using optimization techniques. In Modern Approaches in Machine Learning and Cognitive Science: A Walkthrough: Latest Trends i...

    2021

  46. [54]

    Bishal Lamichhane, Ulf Großekathöfer, Giuseppina Schiavone, and Pierluigi Casale. 2017. Towards stress detection in real-life scenarios using wearable sensors: normalization factor to reduce variability in stress physiology. IneHealth 360°: International Summit on eHealth, Bud...

    2017

  47. [55]

    Remo Lazazzera, Margot Deviaene, Carolina Varon, Bertien Buyse, Dries Testelmans, Pablo Laguna, Eduardo Gil, and Guy Carrault

  48. [56]

    Harim Lee, Eunseon Seong, and Dong-Kyu Chae. 2022. Self-supervised learning with attention-based latent signal augmentation for sleep staging with limited labeled data. In Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence, IJCAI-22, LD R...

    2022

  49. [57]

    Xiang Li, Zhenyan Lu, Dongqi Cai, Xiao Ma, and Mengwei Xu. 2024. Large language models on mobile devices: Measurements, analysis, and insights. In Proceedings of the Workshop on Edge and Mobile Foundation Models . 1–6

    2024

  50. [58]

    Xiaoya Li, Xiaofei Sun, Yuxian Meng, Junjun Liang, Fei Wu, and Jiwei Li. 2019. Dice loss for data-imbalanced NLP tasks. arXiv preprint arXiv:1911.02855 (2019)

    2019 arXiv

  51. [59]

    Yongbo Liang, Zhencheng Chen, Guiyong Liu, and Mohamed Elgendi. 2018. A new, short-recorded photoplethysmogram dataset for blood pressure monitoring in China. Scientific data 5, 1 (2018), 1–7

    2018

  52. [60]

    Hui Lin, Jiyang Li, Ramy Hussein, Xin Sui, Xiaoyu Li, Guangpu Zhu, Aggelos K Katsaggelos, Zijing Zeng, and Yelei Li. 2025. Longitudinal Wrist PPG Analysis for Reliable Hypertension Risk Screening Using Deep Learning. In ICASSP 2025-2025 IEEE International Conference on Acousti...

    2025

  53. [61]

    Guilin Liu, Fitsum A Reda, Kevin J Shih, Ting-Chun Wang, Andrew Tao, and Bryan Catanzaro. 2018. Image inpainting for irregular holes using partial convolutions. In Proceedings of the European conference on computer vision (ECCV) . 85–100

    2018

  54. [62]

    Anna Lo Grasso, Pamela Zontone, Roberto Rinaldo, and Antonio Affanni. 2024. Advanced Necklace for Real-Time PPG Monitoring in Drivers. Sensors 24, 18 (2024), 5908

    2024

  55. [63]

    Cameron McCarthy, Nikhilesh Pradhan, Calum Redpath, and Andy Adler. 2016. Validation of the Empatica E4 wristband. In 2016 IEEE EMBS international student conference (ISC) . IEEE, 1–4

    2016

  56. [64]

    Qianwen Meng, Hangwei Qian, Yong Liu, Yonghui Xu, Zhiqi Shen, and Lizhen Cui. 2023. Unsupervised representation learning for time series: A review. arXiv preprint arXiv:2308.01578 (2023)

    2023 arXiv

  57. [65]

    Steven Miller. 2025. Latest ENSODATA FDA 510(k) clearance enables AI-powered sleep diagnosis using pulse oximetry de- vices. https://www.ensodata.com/press/latest-ensodata-fda-510k-clearance-enables-ai-powered-sleep-diagnosis- using-pulse-oximetry-devices/

    2025

  58. [66]

    Varun Mishra, Tian Hao, Si Sun, Kimberly N Walter, Marion J Ball, Ching-Hua Chen, and Xinxin Zhu. 2018. Investigating the role of context in perceived stress detection in the wild. In Proceedings of the 2018 ACM International Joint Conference and 2018 International Symposium o...

    2018

  59. [67]

    Nikos Mitro, Katerina Argyri, Lampros Pavlopoulos, Dimitrios Kosyvas, Lazaros Karagiannidis, Margarita Kostovasili, Fay Misichroni, Eleftherios Ouzounoglou, and Angelos Amditis. 2023. AI-Enabled Smart Wristband Providing Real-Time Vital Signs and Stress Monitoring. Sensors 23,...

    2023

  60. [68]

    Thierry Moreau, Tianqi Chen, and Luis Ceze. 2018. Leveraging the vta-tvm hardware-software stack for fpga acceleration of 8-bit resnet-18 inference. In Proceedings of the 1st on Reproducible Quality-Efficient Systems Tournament on Co-designing Pareto-efficient Deep Learning. 1

    2018

  61. [69]

    Koorosh Motaman, Khalil Alipour, Bahram Tarvirdizadeh, and Mohammad Ghamari. 2022. A Stress Detection Model Based on LSTM Network Using Solely Raw PPG Signals. In 2022 10th RSI International Conference on Robotics and Mechatronics (ICRoM) . IEEE, 485–490

    2022

  62. [70]

    Sameer Neupane, Mithun Saha, Nasir Ali, Timothy Hnat, Shahin Alan Samiei, Anandatirtha Nandugudi, David M Almeida, and Santosh Kumar. 2024. Momentary Stressor Logging and Reflective Visualizations: Implications for Stress Management with Wearables. arXiv preprint arXiv:2401.16...

    2024 arXiv

  63. [71]

    Maxime Oquab, Timothée Darcet, Théo Moutakanni, Huy Vo, Marc Szafraniec, Vasil Khalidov, Pierre Fernandez, Daniel Haziza, Francisco Massa, Alaaeldin El-Nouby, et al. 2023. Dinov2: Learning robust visual features without supervision. Transactions on Machine Learning Research (2023)

    2023

  64. [72]

    Han Ouyang, Jingjing Tian, Guanglong Sun, Yang Zou, Zhuo Liu, Hu Li, Luming Zhao, Bojing Shi, Yubo Fan, Yifan Fan, et al. 2017. Self-powered pulse sensor for antidiastole of cardiovascular disease. Advanced Materials 29, 40 (2017), 1703456

    2017

  65. [73]

    Yilmazcan Ozyurt, Stefan Feuerriegel, and Ce Zhang. 2022. Contrastive learning for unsupervised domain adaptation of time series. arXiv preprint arXiv:2206.06243 (2022)

    2022 arXiv

  66. [74]

    Arvind Pillai, Dimitris Spathis, Fahim Kawsar, and Mohammad Malekzadeh. 2024. PaPaGei: Open Foundation Models for Optical Physiological Signals. arXiv preprint arXiv:2410.20542 (2024)

    2024 arXiv

  67. [75]

    Kurt Plarre, Andrew Raij, Syed Monowar Hossain, Amin Ahsan Ali, Motohiro Nakajima, Mustafa Al’Absi, Emre Ertin, Thomas Kamarck, Santosh Kumar, Marcia Scott, et al. 2011. Continuous inference of psychological stress from sensory measurements collected in the natural environment...

    2011

  68. [76]

    Ming-Zher Poh, Yukkee Cheung Poh, Pak-Hei Chan, Chun-Ka Wong, Louise Pun, Wangie Wan-Chiu Leung, Yu-Fai Wong, Michelle Man-Ying Wong, Daniel Wai-Sing Chu, and Chung-Wah Siu. 2018. Diagnostic assessment of a deep learning system for detecting atrial Proc. ACM Interact. Mob. Wea...

    2018

  69. [77]

    David Pollreisz and Nima TaheriNejad. 2022. Detection and removal of motion artifacts in PPG signals.Mobile Networks and Applications 27, 2 (2022), 728–738

    2022

  70. [78]

    Jiří Přibil, Anna Přibilová, and Ivan Frollo. 2021. Wearable PPG sensor with bluetooth data transmission for continual measurement in low magnetic field environment. In 2021 International Conference on Applied Electronics (AE) . IEEE, 1–4

    2021

  71. [79]

    Antti Puranen, Tuomas Halkola, Ole Kirkeby, and Antti Vehkaoja. 2020. Effect of skin tone and activity on the performance of wrist-worn optical beat-to-beat heart rate monitoring. In 2020 IEEE SENSORS. IEEE, 1–4

    2020

  72. [80]

    Suha Rabbani and Naimul Khan. 2022. Contrastive self-supervised learning for stress detection from ecg data. Bioengineering 9, 8 (2022), 374

    2022

  73. [81]

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. 2021. Learning transferable visual models from natural language supervision. In International conference on machine learni...

    2021

  74. [82]

    Alec Radford, Karthik Narasimhan, Tim Salimans, Ilya Sutskever, et al . 2018. Improving language understanding by generative pre-training. (2018)

    2018

  75. [83]

    Kevin Rajakariar, Paul Buntine, Andrew Ghaly, Zheng Cheng Zhu, Vihangi Abeygunawardana, Sarah Visakhamoorthy, Patrick J Owen, Shaun Tham, Liam Hackett, Louise Roberts, et al . 2024. Accuracy of Smartwatch Pulse Oximetry Measurements in Hospitalized Patients With Coronavirus Di...

    2024

  76. [85]

    Attila Reiss, Ina Indlekofer, Philip Schmidt, and Kristof Van Laerhoven. 2019. Deep PPG: Large-scale heart rate estimation with convolutional neural networks. Sensors 19, 14 (2019), 3079

    2019

  77. [86]

    Mantas Rinkevičius, Spyridon Kontaxis, Eduardo Gil, Raquel Bailón, Jesús Lázaro, Pablo Laguna, and Vaidotas Marozas. 2019. Photo- plethysmogram signal morphology-based stress assessment. In 2019 Computing in Cardiology (CinC) . IEEE, Page–1

    2019

  78. [87]

    Mostafa Salah, Osama A Omer, Loay Hassan, Mohamed Ragab, Ammar Mostafa Hassan, and Ahmed Abdelreheem. 2022. Beat-based PPG-ABP cleaning technique for blood pressure estimation. IEEE Access 10 (2022), 55616–55626

    2022

  79. [88]

    Nazir Saleheen, Md Azim Ullah, Supriyo Chakraborty, Deniz S Ones, Mani Srivastava, and Santosh Kumar. 2021. Wristprint: Character- izing user re-identification risks from wrist-worn accelerometry data. In Proceedings of the 2021 ACM SIGSAC Conference on Computer and Communicat...

    2021

  80. [89]

    Fatemeh Sarhaddi, Kianoosh Kazemi, Iman Azimi, Rui Cao, Hannakaisa Niela-Vilén, Anna Axelin, Pasi Liljeberg, and Amir M Rahmani

  81. [90]

    Philip Schmidt, Robert Dürichen, Attila Reiss, Kristof Van Laerhoven, and Thomas Plötz. 2019. Multi-target affect detection in the wild: an exploratory study. In Proceedings of the 2019 ACM International Symposium on Wearable Computers . 211–219

    2019

  82. [92]

    Philip Schmidt, Attila Reiss, Robert Duerichen, Claus Marberger, and Kristof Van Laerhoven. 2018. Introducing wesad, a multimodal dataset for wearable stress and affect detection. In Proceedings of the 20th ACM international conference on multimodal interaction . 400–408

    2018

  83. [93]

    Hangsik Shin and Se Dong Min. 2017. Feasibility study for the non-invasive blood pressure estimation based on ppg morphology: Normotensive subject study. Biomedical engineering online 16 (2017), 1–14

    2017

  84. [94]

    Elena Smets, Emmanuel Rios Velazquez, Giuseppina Schiavone, Imen Chakroun, Ellie D’Hondt, Walter De Raedt, Jan Cornelis, Olivier Janssens, Sofie Van Hoecke, Stephan Claes, et al. 2018. Large-scale wearable data reveal digital phenotypes for daily-life stress detection. NPJ dig...

    2018

  85. [95]

    Dimitris Spathis, Ignacio Perez-Pozuelo, Laia Marques-Fernandez, and Cecilia Mascolo. 2022. Breaking away from labels: The promise of self-supervised machine learning in intelligent health. Patterns 3, 2 (2022)

    2022

  86. [96]

    Chi Ian Tang, Ignacio Perez-Pozuelo, Dimitris Spathis, and Cecilia Mascolo. 2020. Exploring contrastive learning in human activity recognition for healthcare. arXiv preprint arXiv:2011.11542 (2020)

    2020 arXiv

  87. [97]

    Sana Tonekaboni, Danny Eytan, and Anna Goldenberg. 2021. Unsupervised representation learning for time series with temporal neighborhood coding. International Conference of Learning Representations (2021)

    2021

  88. [98]

    Kobiljon Toshnazarov, Uichin Lee, Byung Hyung Kim, Varun Mishra, Lismer Andres Caceres Najarro, and Youngtae Noh. 2024. SOSW: Stress Sensing with Off-the-shelf Smartwatches in the Wild. IEEE Internet of Things Journal (2024)

    2024

  89. [99]

    James Truslow, Angela Spillane, Huiming Lin, Katherine Cyr, Adeeti Ullal, Edith Arnold, Ron Huang, Laura Rhodes, Jennifer Block, Jamie Stark, et al. 2024. Understanding activity and physiology at scale: The Apple Heart & Movement Study. npj Digital Medicine 7, 1 (2024), 242. P...

    2024

  90. [100]

    Merel M van Gilst, Johannes P van Dijk, Roy Krijn, Bertram Hoondert, Pedro Fonseca, Ruud JG van Sloun, Bruno Arsenali, Nele Vandenbussche, Sigrid Pillen, Henning Maass, et al . 2019. Protocol of the SOMNIA project: an observational study to create a neurophysiological database...

    2019

  91. [101]

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

    2017

  92. [102]

    Jie Wang, Tuantuan Lu, Ruogu Huang, and Yongxiang Zhao. 2024. Classifying engagement in E-learning through GRU-TCN model using photoplethysmography signals. Biomedical Signal Processing and Control 90 (2024), 105903

    2024

  93. [103]

    Stephanie J Wilson, Brittney E Bailey, Lisa M Jaremka, Christopher P Fagundes, Rebecca Andridge, William B Malarkey, Kathleen M Gates, and Janice K Kiecolt-Glaser. 2018. When couples’ hearts beat together: Synchrony in heart rate variability during conflict predicts heightened...

    2018

  94. [104]

    Gerald Woo, Chenghao Liu, Doyen Sahoo, Akshat Kumar, and Steven Hoi. 2022. CoST: Contrastive learning of disentangled seasonal- trend representations for time series forecasting. arXiv preprint arXiv:2202.01575 (2022)

    2022 arXiv

  95. [105]

    Maxwell Xu, Alexander Moreno, Supriya Nagesh, Varol Aydemir, David Wetter, Santosh Kumar, and James M Rehg. 2022. PulseImpute: A Novel Benchmark Task for Pulsative Physiological Signal Imputation. Advances in Neural Information Processing Systems Dataset and Benchmarks Track 3...

    2022

  96. [107]

    Maxwell Xu, Alexander Moreno, Hui Wei, Benjamin Marlin, and James Matthew Rehg. 2024. REBAR: Retrieval-Based Reconstruction for Time-series Contrastive Learning. In The Twelfth International Conference on Learning Representations

    2024

  97. [108]

    Maxwell A Xu, Jaya Narain, Gregory Darnell, Haraldur Hallgrimsson, Hyewon Jeong, Darren Forde, Richard Fineman, Karthik J Raghuram, James M Rehg, and Shirley Ren. 2024. RelCon: Relative Contrastive Learning for a Motion Foundation Model for Wearable Data. arXiv preprint arXiv:...

    2024 arXiv

  98. [109]

    Jiamei Yang, Yu Wang, Hui Wang, Peng Zhou, Xinyou Li, and Jianbin Zheng. 2024. Implementation of FPGA-based ResNet accelerator for vehicle detection. In Ninth International Symposium on Advances in Electrical, Electronics, and Computer Engineering (ISAEECE 2024) , Vol. 13291. ...

    2024

  99. [110]

    Ling Yang and Shenda Hong. 2022. Unsupervised time-series representation learning with iterative bilinear temporal-spectral fusion. In International Conference on Machine Learning . PMLR, 25038–25054

    2022

  100. [111]

    Xinyu Yang, Zhenguo Zhang, and Rongyi Cui. 2022. Timeclr: A self-supervised contrastive learning framework for univariate time series representation. Knowledge-Based Systems 245 (2022), 108606

    2022

  101. [112]

    Bowen Yao, Liansheng Liu, Yu Peng, and Xiyuan Peng. 2023. Intelligent measurement on edge devices using hardware memory-aware joint compression enabled neural networks. IEEE Transactions on Instrumentation and Measurement 73 (2023), 1–13

    2023

  102. [113]

    Zhihan Yue, Yujing Wang, Juanyong Duan, Tianmeng Yang, Congrui Huang, Yunhai Tong, and Bixiong Xu. 2022. Ts2vec: Towards universal representation of time series. In Proceedings of the AAAI Conference on Artificial Intelligence , Vol. 36. 8980–8987

    2022

  103. [114]

    Taedong Yun, Justin Cosentino, Babak Behsaz, Zachary R McCaw, Davin Hill, Robert Luben, Dongbing Lai, John Bates, Howard Yang, Tae-Hwi Schwantes-An, et al. 2024. Unsupervised representation learning on high-dimensional clinical data improves genomic discovery and prediction. N...

    2024

  104. [115]

    Panyu Zhang, Gyuwon Jung, Jumabek Alikhanov, Uzair Ahmed, and Uichin Lee. 2024. A Reproducible Stress Prediction Pipeline with Mobile Sensor Data. Proceedings of the ACM on interactive, mobile, wearable and ubiquitous technologies 8, 3 (2024), 1–35

    2024

  105. [116]

    Xiang Zhang, Ziyuan Zhao, Theodoros Tsiligkaridis, and Marinka Zitnik. 2022. Self-supervised contrastive pre-training for time series via time-frequency consistency. Advances in Neural Information Processing Systems 35 (2022), 3988–4003

    2022

  106. [117]

    Zexing Zhang, Huimin Lu, Songzhe Ma, Jianzhong Peng, Chenglin Lin, Niya Li, and Bingwang Dong. 2024. A general framework for generative self-supervised learning in non-invasive estimation of physiological parameters using photoplethysmography. Biomedical Signal Processing and ...

    2024

  107. [118]

    Ce Zhou, Qian Li, Chen Li, Jun Yu, Yixin Liu, Guangjing Wang, Kai Zhang, Cheng Ji, Qiben Yan, Lifang He, et al. 2024. A comprehensive survey on pretrained foundation models: A history from bert to chatgpt. International Journal of Machine Learning and Cybernetics (2024), 1–65

    2024

  108. [119]

    Lili Zhu, Petros Spachos, Pai Chet Ng, Yuanhao Yu, Yang Wang, Konstantinos Plataniotis, and Dimitrios Hatzinakos. 2023. Stress detection through wrist-based electrodermal activity monitoring and machine learning. IEEE Journal of Biomedical and Health Informatics 27, 5 (2023), ...

    2023

  109. [2019]

    NPJ digital medicine 2, 1 (2019), 60

    The use of photoplethysmography for assessing hypertension. NPJ digital medicine 2, 1 (2019), 60

    2019

  110. [2020]

    IEEE Transactions on Biomedical Engineering 68, 5 (2020), 1496–1506

    Detection and classification of sleep apnea and hypopnea using PPG and SpO _2 signals. IEEE Transactions on Biomedical Engineering 68, 5 (2020), 1496–1506

    2020

  111. [2022]

    PloS one 17, 12 (2022), e0268361

    A comprehensive accuracy assessment of Samsung smartwatch heart rate and heart rate variability. PloS one 17, 12 (2022), e0268361

    2022

  112. [2023]

    arXiv preprint arXiv:2312.10308 (2023)

    Event-Based Contrastive Learning for Medical Time Series. arXiv preprint arXiv:2312.10308 (2023)

    2023 arXiv

  113. [2024]

    IEEE Transactions on Pattern Analysis and Machine Intelligence (2024)

    SpectralGPT: Spectral remote sensing foundation model. IEEE Transactions on Pattern Analysis and Machine Intelligence (2024)

    2024

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.