Pith. sign in

REVIEW 4 major objections 6 minor 25 references

Sedentary Behavior Classification for Wearable Sensors with a CNN-BiLSTM Model

T0 review · 4 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read Hip-pretrained CHAP transfers to wrist accelerometer data after finetuning, beating a from-scratch transformer on window-level balanced accuracy.

desk verdict Useful hip-to-wrist transfer study undermined by an internal contradiction between the abstract's 'consistent advantages' claim and the paper's own subject-level tables; the low-label data-efficiency finding holds up. read the letter →

arxiv 2608.02946 v1 pith:KWCKRCTJ submitted 2026-08-03 cs.LG

Yuliang Chen , Weiwei Shi , Jingjing Zou , Rong Zablocki , Animesh Kumar , Jordan A. Carlson , Sheri J. Hartman , Mikael Anne Greenwood-Hickman
show 5 more authors
Paul R. Hibbing Marta Jankowska Jay Yang Arun Kumar Loki Natarajan
This is my paper · ORCID
classification cs.LG
keywords sedentarybehaviorwearableaccelerometerstransferlearningCNN-BiLSTMwrist-wornsensorship-wornsittingclassificationposturerecognition
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

This paper asks whether a deep-learning posture classifier trained on hip-worn accelerometers can be reused on wrist-worn accelerometers, where sensors are more comfortable and more likely to be worn continuously. On the iWatch cohort, the hip-pretrained CHAP model keeps its accuracy on new hip data with no retraining, but its balanced accuracy drops from 88.7% to 71.8% when moved to the wrist. Finetuning on labeled wrist data raises wrist balanced accuracy to 82.6%, outperforming a transformer trained from scratch on the same data (79.4%) and needing far fewer labels to get there. The paper concludes that hip-based pretraining is a useful starting point for wrist deployment, but that wrist-specific adaptation is required because hip motion is more tightly coupled to posture than wrist motion.

What carries the argument

The load-bearing object is CHAP, a CNN-BiLSTM sequence model. A CNN encodes each non-overlapping 10-second triaxial acceleration window into a 128-dimensional feature vector, a bidirectional LSTM pools temporal context across the full 7-minute input segment, and a linear head emits a per-window sitting probability. The pretrained weights embody hip-specific signal statistics; comparing CHAP with pretrained weights against the same architecture initialized randomly isolates what the pretraining contributes. A secondary mechanism is the distribution-shift quantifier, the Jensen-Shannon distance (0.486) between hip and wrist window statistics, which motivates finetuning. The finetuning procedure itself, updating all parameters on target-placement data under a class-weighted cross-entropy loss, is what converts the hip representation into a wrist-usable classifier.

What would settle it

Re-run the wrist finetuning comparison with ground truth from a thigh-worn activPAL device on the same cohort, and check whether finetuned CHAP still beats a from-scratch transformer by about three balanced-accuracy points; if the gap narrows or reverses, the SenseCam collapsed labels, particularly the sitting and vehicle grouping, were carrying the result.

Watch

Extended reading notes

Core claim

The central claim is that cross-placement transfer for sedentary-behavior classification is partially possible: what transfers is not the raw decision boundary but the learned representation. CHAP, a CNN-BiLSTM trained on hip data from 1,397 adults with thigh-worn activPAL labels, achieves 88.74% balanced accuracy on the iWatch hip test set with no finetuning, confirming within-placement generalization. On wrist data the same zero-shot model falls to 71.83%, and the hip-versus-wrist signal distributions are widely separated (Jensen-Shannon distance 0.486). Supervised finetuning on the iWatch wrist training set recovers most of the gap, reaching 82.56% balanced accuracy, which is 3.2 points above a ViT-Small transformer trained from scratch on identical data, and the finetuned CHAP also shows markedly better data efficiency at 1%, 10%, and 50% label budgets. The paper also documents a wrist-specific error asymmetry: after finetuning, 22.9% of true non-sitting windows are labeled sitting, reflecting low-motion upright activities whose wrist signals resemble sitting; a similar bias is larger for the transformer (30.7%) and for zero-shot CHAP on wrist. The authors conclude that hip pretraining helps mainly through faster convergence and lower label requirements, not by eliminating the placement gap.

Load-bearing premise

The ground-truth posture labels come from wearable-camera images captured roughly every 20 seconds and collapsed into sitting (Sedentary, Vehicle) and non-sitting (Standing Still, Standing Moving, Walking/Running); if those collapsed labels are systematically wrong, especially grouping Vehicle with Sitting, every accuracy figure and transfer conclusion inherits the error.

Editorial extensions

If this is right

  • Hip-pretrained CHAP is a practical initialization for wrist deployment: finetuning with 10% of labeled wrist data already yields most of the benefit, so new studies can avoid collecting and annotating large wrist datasets.
  • Under matched training conditions, finetuned CHAP beats a from-scratch transformer at window-level balanced accuracy despite the transformer having more parameters, indicating that inductive bias from CNN-BiLSTM matters when labels are limited.
  • The wrist-specific error pattern, low-motion non-sitting labeled as sitting, persists after finetuning, so wrist-based estimates of sitting breaks and bout duration will tend to overestimate sitting time and underestimate breaks unless corrected.
  • With sufficient wrist labels and training time, random initialization approaches finetuned performance, so the pretraining benefit is chiefly data efficiency and early-epoch stability rather than an asymptotic accuracy ceiling.
  • On hip data, zero-shot CHAP already performs near saturation, so hip-to-hip transfer needs no adaptation; finetuning adds little.

Reading between the lines

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

  • The paper compares models on a single cohort; a direct extension would test whether the same finetuning recipe transfers across cohorts and device brands, since the iWatch data come from one ActiGraph model and one population.
  • The mixed subject-level results, where ViT Small produced lower error for total sedentary time while CHAP FT had better balanced accuracy, suggest the best model depends on the downstream metric; a metric-aware finetuning objective could combine both strengths.
  • Grouping Vehicle with Sitting and Standing Moving with Non-sitting makes the binary task easier than real posture inference; re-annotating those categories separately would show how much of the reported accuracy is due to this collapse.
  • The data-efficiency curves imply a practical recipe: collect roughly 10% of a target cohort's wrist labels and finetune hip-pretrained CHAP, which would cut annotation cost substantially in epidemiology studies.
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

4 major / 6 minor

Summary. The paper evaluates whether CHAP, a CNN-BiLSTM model pretrained on hip-worn accelerometer data for sitting versus non-sitting classification, transfers to wrist-worn accelerometer data. Using the iWatch dataset with SenseCam-derived posture labels, the authors compare zero-shot CHAP, fine-tuned CHAP (CHAP FT), a randomly initialized CHAP, and a ViT Small transformer trained from scratch. They report window-level balanced accuracy/F1, subject-level clinical metrics (total sedentary time, sit-to-stand transitions, bout durations), and label-efficiency curves at 1%, 10%, 50%, and 100% of the training data. The main claimed findings are that hip pretraining provides a useful starting point for wrist adaptation, that fine-tuning improves wrist performance and data efficiency, and that fine-tuned CHAP has 'consistent advantages' over transformers trained from scratch.

Significance. If the transfer result is robust, the study addresses a practically important question: whether the large existing investment in hip-worn CHAP models can be leveraged for wrist-worn devices, which are more comfortable and more common in consumer wearables. The study uses a large free-living dataset with participant-level splitting, and the limited-label analysis (Figure 4 and Table VII) is a valuable design for quantifying annotation cost. The sensitivity analysis comparing pretrained and random initialization is also informative. However, the headline comparative claim against transformers is not supported by the paper's own subject-level tables, and several numerical inconsistencies call the reliability of the reported results into question. The practical transfer value of hip pretraining for wrist deployment is moderately supported, but the manuscript needs substantial revision before its central claims can be accepted.

major comments (4)
  1. [Abstract; Section III.C; Tables III and V] The abstract states that 'Finetuning CHAP provides consistent advantages over transformer models trained from scratch,' and Section III.C states that on wrist, 'CHAP FT clearly outperformed both CHAPZS and ViT Small' with CHAP FT exceeding ViT Small by approximately 7% in F1. These claims are directly contradicted by the paper's own results. In Table III (wrist), CHAP FT has F1 = 0.80 while ViT Small has F1 = 0.87, i.e., ViT exceeds CHAP FT by 7 points. In Table V, ViT Small achieves the lower MAPE on all four wrist subject-level sedentary metrics: total sedentary time 10.9% vs. 14.2%, sit-to-stand transitions/day 31.5% vs. 61.3%, time in bouts ≥30 min 15.0% vs. 20.6%, and mean bout duration 27.1% vs. 42.5%. The 'consistent advantages' claim must be removed or qualified to the specific settings where CHAP FT does win, namely window-level balanced accuracy (82.56% vs. 79.39%) and low-label validation accuracy. This is not a wording issue; it is the paper's central comparative conclusion.
  2. [Section IV.B; Tables VI and VII] There is a numerical inconsistency between the main sensitivity analysis and the limited-label experiment for the randomly initialized CHAP at 100% hip training data. Table VI reports CHAP Random Init hip validation balanced accuracy of 91.22% and test balanced accuracy of 88.38%, while Table VII reports 92.98% validation and 90.35% test for the same configuration. The difference (90.35% vs. 88.38% test) is material: if Table VII is correct, the randomly initialized model outperforms CHAP FT (88.71%) on the hip test set, which would weaken the claim that 'both models reach similar balanced accuracy after sufficient training' and would change the interpretation of the convergence shown in Figure 3. The authors should reconcile these two tables and specify which result corresponds to the configuration used for the rest of the paper.
  3. [Section II.A; Section III] The ground-truth labels are derived from SenseCam images collapsed into Sitting (Sedentary, Vehicle) and Non-sitting (Standing Still, Standing Moving, Walking/Running), and the paper cites [12] for inter-rater reliability but does not report the reliability statistic for the collapsed binary classes. Because every reported accuracy, MAPE, and transfer conclusion inherits any systematic annotation error, the manuscript should either report the inter-rater reliability (kappa or percent agreement) for the binary collapse, or provide a sensitivity analysis that excludes or reclassifies the Vehicle and Standing Moving categories. Without this, the reader cannot judge how much of the hip-to-wrist gap and the model differences is due to label noise rather than sensor placement or model architecture.
  4. [Section III.C] The sentence 'CHAP FT exceeded ViTSmall by approximately 7%' in the subject-level discussion is an internal misreading of Table III: in the wrist row, ViT Small's F1 of 0.87 exceeds CHAP FT's F1 of 0.80 by approximately 7 points, not the reverse. This error also propagates to the claim that 'CHAP FT clearly outperformed both CHAPZS and ViT Small.' The text needs to be corrected to state which metrics favor which model, with the observed differences (e.g., specificity 0.75 vs. 0.68 favors CHAP FT; F1 0.80 vs. 0.87 favors ViT Small).
minor comments (6)
  1. [Figure 5; Section V] The figure caption refers to subject i067A while the text mentions i0167A; the identifiers should be made consistent.
  2. [Table VII] The heading 'Valid acc. Test acc.' does not reflect the train/validation/test structure used in Table VI; clarify whether the reported numbers are validation and test only, and why Table VII has no training column.
  3. [Table II] Each entry in the confusion matrices is a row-normalized fraction of true class, but this is not stated in the caption or table; add a note that rows sum to 1.
  4. [Table VIII] The classifier output size is listed as '42×11', which appears to be a typo for '42×1' or '1'; please correct the dimension.
  5. [Section VII] The word 'setings' in the second paragraph should be 'settings'.
  6. [Section II.B] The paper assumes the CHAP pretrained weights are available from [8], but it does not state where they can be obtained or whether they are public; for reproducibility, please provide a link or a clear statement of availability.

Circularity Check

0 steps flagged · score 1.0 of 10

No material circularity: the transfer claims are tested directly on held-out iWatch data; CHAP self-citation is contextual, not load-bearing.

full rationale

The paper is an empirical evaluation rather than a derivation from first principles. CHAP is the authors' prior hip model, but every central claim (hip-to-wrist zero-shot drop, finetuning gains, data efficiency, comparison with ViT) is measured on a held-out iWatch test set with participant-level splitting, not derived from the CHAP paper [8]. The pretrained weights from [8] are an input, but the conclusions are not equivalent to that input: CHAP FT is compared against CHAP with random initialization under matched training conditions (Table VI, Fig. 3), and CHAP ZS is evaluated without retraining (Table I). The limited-label analysis (Fig. 4, Table VII) directly tests whether pretraining helps with fewer labels. No fitted quantity is relabeled as a prediction; class weights, augmentation, and training hyperparameters are disclosed. The only self-citation is the use of the CHAP model itself, whose hip validation in [8] is an independent external benchmark against activPAL, and the paper re-measures hip zero-shot performance here (88.74% balanced accuracy), so the argument does not reduce to the citation. The abstract's 'consistent advantages' wording is internally contradicted by subject-level tables (e.g., Table V shows ViT Small with lower MAPE on all four wrist metrics), but that is a correctness and claims-scope issue, not circularity. A minor nonzero score reflects the self-citation of CHAP, which is not load-bearing.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The central claim rests on the quality of the SenseCam-derived ground truth, the validity of the CHAP pretrained weights from prior work, and standard preprocessing tools. The only hand-tuned free parameters are the loss class weight and standard training hyperparameters, neither of which is fitted to the test set.

free parameters (2)
  • Class weight for minority class in loss = 2.80 (hip), 2.82 (wrist)
    Weighted binary cross-entropy loss; the paper reports it improved validation balanced accuracy from 92.15% to 92.29% on hip and 84.23% to 86.01% on wrist, so it is a tuned hyperparameter affecting results.
  • Training hyperparameters = lr=1e-3, weight decay=1e-3, batch=128, 40 epochs, 8-epoch warmup
    Chosen by hand per standard practice; not stated to be tuned per placement, but they shape the finetuning curves.
assumptions (3)
  • domain assumption SenseCam image annotations provide valid ground truth for sitting vs non-sitting posture.
    Labels are derived from images captured every ~20 seconds and annotated with an inter-rater reliability protocol [12], but the reliability statistic is not reported in this paper.
  • domain assumption The CHAP model's hip-pretrained weights, trained on 1,397 adults with activPAL labels in [8], are a reasonable feature initialization for the iWatch cohort.
    The paper relies on this pretrained model as the transfer source without re-evaluating its original validation in this paper.
  • domain assumption Choi algorithm correctly removes nonwear periods.
    Nonwear detection is standard, but any misclassification would affect window composition and subsequent metrics.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Sedentary Behavior Classification for Wearable Sensors with a CNN-BiLSTM Model." pith.science (2026). https://pith.science/paper/KWCKRCTJ

@misc{pith2026260802946,
  author       = {Pith},
  title        = {Pith review of: Sedentary Behavior Classification for Wearable Sensors with a CNN-BiLSTM Model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KWCKRCTJ}},
  note         = {Machine review of arXiv:2608.02946}
}
read the original abstract

Accurate detection of sedentary behavior is important for studying health risks related to prolonged sitting, but posture-based classification remains challenging with wearable sensors, especially at the wrist. We study whether a deep learning model trained on hip-worn accelerometer data can transfer to wrist-worn accelerometer data for sitting versus non-sitting classification. We use CHAP, a CNN-BiLSTM model originally developed for hip accelerometers, and evaluate its zero-shot performance on wrist data as well as its adaptation through finetuning with varying amounts of labeled wrist data. Experiments are conducted on the iWatch dataset with ground-truth posture labels derived from wearable cameras. The hip-trained model performs strongly on hip data without retraining, but accuracy drops on wrist data due to sensor placement shift. Finetuning CHAP provides consistent advantages over transformer models trained from scratch. These findings suggest that hip-based pretraining provides a useful starting point for wrist deployment, while highlighting the need for wrist-specific adaptation to handle higher signal variability.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

25 extracted references · 19 canonical work pages

  1. [12]

    Using the SenseCam to improve classifications of sedentary behavior in free-living settings,

    J. Kerr, S. J. Marshall, S. Godbole, J. Chen, A. Legge, A. R. Doherty, P. Kelly, M. Oliver, H. M. Badland, and C. Foster, “Using the SenseCam to improve classifications of sedentary behavior in free-living settings,” American Journal of Preventive Medicine, vol. 44, no. 3, pp. 290–296, Mar. 2013

    work page 2013

  2. [1]

    Sedentary time and its association with risk for disease incidence, mortality, and hospitalization in adults: A systematic review and meta-analysis,

    A. Biswas, P. I. Oh, G. E. Faulkner, R. R. Bajaj, M. A. Silver, M. S. Mitchell, and D. A. Alter, “Sedentary time and its association with risk for disease incidence, mortality, and hospitalization in adults: A systematic review and meta-analysis,”Annals of Internal Medicine, vol. 162, no. 2, pp. 123–132, Jan. 2015, erratum in: Ann Intern Med. 2015 Sep 1;1...

    work page doi:10.7326/l15-5134 2015

  3. [2]

    Sit less and move more for cardiovascular health: emerging insights and opportunities,

    D. W. Dunstan, S. Dogra, S. E. Carter, and N. Owen, “Sit less and move more for cardiovascular health: emerging insights and opportunities,” Nature Reviews Cardiology, vol. 18, no. 9, pp. 637–648, 2021, epub 2021 May 20

    work page 2021

  4. [3]

    Breaks in sedentary time: beneficial associations with metabolic risk,

    G. N. Healy, D. W. Dunstan, J. Salmon, E. Cerin, J. E. Shaw, P. Z. Zimmet, and N. Owen, “Breaks in sedentary time: beneficial associations with metabolic risk,”Diabetes Care, vol. 31, no. 4, pp. 661–666, Apr. 2008, epub 2008 Feb 5

    work page 2008

  5. [4]

    V . T. van Hees, J. H. Migueles, S. Sabia, M. R. Patterson, Z. Fang, J. Heywood, J. C. Pujol, L. Kushleyeva, M. Chen, M. Yerramalla, P. Bos, T. Sanders, C. Zhao, I. M. Danilevicz, V . B. Mesquita, G. Segantin, Medical Research Council UK, Accelting, and French National Research Agency,GGIR: Raw Accelerometer Data Analysis, 2025, r package version 3.3-0. [...

    work page arXiv 2025

  6. [5]

    Agreement of sedentary behaviour metrics derived from hip worn and thigh worn accelerometers among older adults with implications for studying physical and cognitive health,

    J. Bellettiere, F. Tuz Zahra, J. A. Carlson, N. D. Ridgers, S. Liles, M. A. Greenwood Hickman, R. L. Walker, A. Z. LaCroix, M. M. Jankowska, D. E. Rosenberg, and L. Natarajan, “Agreement of sedentary behaviour metrics derived from hip worn and thigh worn accelerometers among older adults with implications for studying physical and cognitive health,”Journa...

    work page 2021

  7. [6]

    Physical activity recognition using posterior-adapted class-based fusion of multiaccelerometer data,

    A. K. Chowdhury, D. Tjondronegoro, V . Chandran, and S. G. Trost, “Physical activity recognition using posterior-adapted class-based fusion of multiaccelerometer data,”IEEE Journal of Biomedical and Health Informatics, vol. 22, no. 3, pp. 678–685, May 2018

    work page 2018

  8. [7]

    Activity recognition using a single accelerometer placed at the wrist or ankle,

    A. Mannini, S. S. Intille, M. Rosenberger, A. M. Sabatini, and W. Haskell, “Activity recognition using a single accelerometer placed at the wrist or ankle,”Medicine and Science in Sports and Exercise, vol. 45, no. 11, pp. 2193–2203, 2013

    work page 2013

Show all 25 references
  1. [8]

    The CNN Hip Accelerometer Posture (CHAP) method for classifying sitting patterns from hip accelerometers: A validation study,

    M. A. Greenwood-Hickman, S. Nakandala, M. M. Jankowska, D. E. Rosenberg, F. Tuz-Zahra, J. Bellettiere, J. Carlson, P. R. Hibbing, J. Zou, A. Z. LaCroix, A. Kumar, and L. Natarajan, “The CNN Hip Accelerometer Posture (CHAP) method for classifying sitting patterns from hip accel...

    2021

  2. [9]

    Accelerometry data in health research: challenges and opportunities,

    M. Karas, J. Bai, M. Str ˛ aczkiewicz, J. Harezlak, N. W. Glynn, T. Harris, V . Zipunnikov, C. Crainiceanu, and J. K. Urbanek, “Accelerometry data in health research: challenges and opportunities,”Statistical Biosciences, vol. 11, no. 2, pp. 210–237, Jul. 2019, epub 2019 Jan 12

    2019

  3. [10]

    Transfer learning improves accelerometer-based child activity recognition via subject-independent adult-domain adaption,

    J. Li, P. Kang, T. Tan, and P. Shull, “Transfer learning improves accelerometer-based child activity recognition via subject-independent adult-domain adaption,”IEEE Journal of Biomedical and Health Infor- matics, vol. PP, p. 1, 10 2021

    2021

  4. [11]

    Automatically assisting human memory: a SenseCam browser,

    A. R. Doherty, C. J. A. Moulin, and A. F. Smeaton, “Automatically assisting human memory: a SenseCam browser,”Memory, vol. 19, no. 7, pp. 785–795, Oct. 2011

    2011

  5. [13]

    Using accelerometers in youth physical activity studies: a review of methods,

    K. L. Cain, J. F. Sallis, T. L. Conway, D. Van Dyck, and L. Calhoon, “Using accelerometers in youth physical activity studies: a review of methods,”Journal of Physical Activity and Health, vol. 10, no. 3, pp. 437–450, Mar. 2013

    2013

  6. [14]

    Best practices for using physical activity monitors in population based research,

    C. E. Matthews, M. Hagströmer, D. M. Pober, and H. R. Bowles, “Best practices for using physical activity monitors in population based research,”Medicine and Science in Sports and Exercise, vol. 44, no. 1 Suppl 1, pp. S68–S76, Jan. 2012

    2012

  7. [15]

    Application of convolutional neural network algorithms for advancing sedentary and activity bout classification,

    S. Nakandala, M. M. Jankowska, F. Tuz-Zahra, J. Bellettiere, J. A. Carlson, A. Z. LaCroix, S. J. Hartman, D. E. Rosenberg, J. Zou, A. Kumar, and L. Natarajan, “Application of convolutional neural network algorithms for advancing sedentary and activity bout classification,”Jour...

    2021

  8. [16]

    Validation of accelerometer wear and nonwear time classification algorithm,

    L. Choi, Z. Liu, C. E. Matthews, and M. S. Buchowski, “Validation of accelerometer wear and nonwear time classification algorithm,” Medicine and Science in Sports and Exercise, vol. 43, no. 2, pp. 357– 364, Feb 2011

    2011

  9. [17]

    Attention Is All You Need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, L. Kaiser, and I. Polosukhin, “Attention Is All You Need,”CoRR, vol. abs/1706.03762, 2017. [Online]. Available: http://arxiv.org/abs/1706.03762

    2017 arXiv

  10. [18]

    Sundial: A family of highly capable time series foundation models,

    Y . Liu, G. Qin, Z. Shi, Z. Chen, C. Yang, X. Huang, J. Wang, and M. Long, “Sundial: A family of highly capable time series foundation models,” 2025. [Online]. Available: https://arxiv.org/abs/2502.00816

    2025 arXiv

  11. [19]

    A time series is worth 64 words: Long-term forecasting with Transformers,

    Y . Nie, N. H. Nguyen, P. Sinthong, and J. Kalagnanam, “A time series is worth 64 words: Long-term forecasting with Transformers,” 2023. [Online]. Available: https://arxiv.org/abs/2211.14730

    2023 arXiv

  12. [20]

    An image is worth 16x16 words: Transformers for Image Recognition at scale,

    A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly, J. Uszkoreit, and N. Houlsby, “An image is worth 16x16 words: Transformers for Image Recognition at scale,” 2021. [Online]. Available: https://arxi...

    2021 arXiv

  13. [21]

    Divergence measures based on the Shannon entropy,

    J. Lin, “Divergence measures based on the Shannon entropy,”IEEE Transactions on Information Theory, vol. 37, no. 1, pp. 145–151, 1991

    1991

  14. [22]

    Amount of time spent in sedentary behaviors in the United States, 2003-2004

    C. Matthews, K. Chen, P. Freedson, M. Buchowski, B. Beech, R. Pate, and R. Troiano, “Amount of time spent in sedentary behaviors in the United States, 2003-2004.”Am J Epidemiol., vol. 167, no. 7, pp. 875–81, Apr 2008

    2003

  15. [23]

    Comparison of sedentary estimates between activPAL and hip- and wrist-worn ActiGraph,

    A. Koster, E. Shiroma, P. Caserotti, C. Matthews, K. Chen, N. Glynn, and T. Harris, “Comparison of sedentary estimates between activPAL and hip- and wrist-worn ActiGraph,”Med Sci Sports Exerc., vol. 48, no. 8, pp. 1514–1522, Aug 2016

    2016

  16. [24]

    Scaling Vision Transformers,

    X. Zhai, A. Kolesnikov, N. Houlsby, and L. Beyer, “Scaling Vision Transformers,” 2022. [Online]. Available: https://arxiv.org/abs/2106. 04560

    2022

  17. [25]

    Self-supervised learning for human activity recognition using 700,000 person-days of wearable data,

    H. Yuan, S. Chan, A. P. Creagh, C. Tong, A. Acquah, D. A. Clifton, and A. Doherty, “Self-supervised learning for human activity recognition using 700,000 person-days of wearable data,” npj Digital Medicine, vol. 7, no. 1, Apr. 2024. [Online]. Available: http://dx.doi.org/10.10...

    2024 doi

Pith tools

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