Pith. sign in

REVIEW 4 major objections 5 minor 2 cited by

Beyond Confusion: A Fine-grained Dialectical Examination of Human Activity Recognition Benchmark Datasets

T0 review · 4 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Across six widely used human-activity-recognition benchmark datasets, the paper identifies sensor windows that none of six state-of-the-art models can classify, and traces them to dataset-inherent problems rather than model limitations.

desk verdict A solid multi-dataset HAR audit with a useful mask artifact, but the 'inherently ambiguous data' claim outruns the evidence; reframe the IFC as ensemble-relative and release the mask. read the letter →

arxiv 2412.09037 v1 pith:AHAHGUMC submitted 2024-12-12 cs.LG

Daniel Geissler , Dominique Nshimyimana , Vitor Fortes Rey , Sungho Suh , Bo Zhou , Paul Lukowicz This is my paper · ORCID
classification cs.LG
keywords humanactivityrecognitiondatasetqualitylabelambiguityintersectoffalseclassificationstrinarymaskwearablesensorsbenchmarkdatasetscross-validation
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 tries to establish that on six widely used human-activity-recognition (HAR) benchmark datasets, a measurable share of sensor windows is mislabeled or ambiguous in a way that no classifier can be expected to learn, and that this data-level ceiling, not model expressiveness, is what keeps reported accuracies below near-perfect levels. The authors train six milestone model architectures under grouped leave-out cross-validation, intersect the windows every model gets wrong into the Intersect of False Classifications (IFC), and then inspect those windows at signal level to identify the causes: ambiguous annotations, recording execution irregularities, and misaligned transition periods. For a reader, the payoff is a concrete way to separate 'model failure' from 'data failure' in HAR benchmarks, plus a trinary mask that marks each window clean, minor, or major so datasets can be patched before being reused.

What carries the argument

The load-bearing object is the Intersect of False Classifications (IFC): the set of sliding windows that none of the six models in the panel classifies correctly, computed after training each model under grouped leave-out cross-validation and merging the false classifications across models. It is defined as $\mathrm{IFC} = 100\% - \text{common ground} - \sum_i \text{single contribution}_i$, where common ground is the share of windows all models classify correctly and single contributions are windows only one model gets right. The IFC converts a set of model errors into a dataset property, and its subdivision into a trinary mask carries the practical argument: the mask labels each window clean, minor, or major by taking the gap between the highest predicted probability and the next highest, so confidently wrong prediction windows are marked major and uncertain ones minor.

What would settle it

Take the IFC windows from any of the six datasets and train a substantially different classifier, one with a different window size or stride, a model that sees surrounding context, or an oracle that uses synchronized video of the recording session; if any such system correctly labels a large share of IFC windows, the inference from model failure to dataset ambiguity is refuted. A second check: have independent human annotators re-label the IFC segments without knowing the original labels; if they consistently agree on a single class, those segments are not inherently ambiguous.

Watch

Extended reading notes

Core claim

The paper's central claim is that the IFC segments exist in every one of the six datasets and are caused by properties of the dataset itself rather than by weak models: PAMAP2 has an IFC share of 11.74%, Opportunity gesture labels 12.06%, Opportunity locomotion 5.31%, MHealth 4.58%, WISDM 1.61%, while MM-Fit and MotionSense stay below 0.5%. Signal-level inspection attributes the persistent errors to annotation design where classes are not mutually exclusive (e.g., ironing containing standing), to null classes that absorb heterogeneous movements, to recording irregularities such as loose sensors or periodic noise, and to transitions between activities being labeled with too coarse a granularity. The paper concludes that benchmark accuracy on these datasets is capped by data quality, and offers the trinary mask as a dataset patch and a set of collection guidelines as a preventive measure.

Load-bearing premise

The argument collapses if the six chosen models share enough inductive biases that their joint failure says nothing about other classifiers; the paper itself concedes in Section 7.2 that it could not establish a clear measure that this set of models sufficiently represents an appropriate intersection.

Editorial extensions

If this is right

  • Reported accuracies on these benchmarks should be read as ceilings set partly by data quality: for PAMAP2, 11.74% of windows are in the IFC and cannot be learned by any of the tested models.
  • Practitioners can use the trinary mask to filter major segments before training, or to integrate the mask labels into the loss function so models become aware of demanding sections.
  • The contrast between datasets with fine-grained labels (MotionSense, MM-Fit) and those with coarse protocol-based labels (PAMAP2) indicates that annotation design, not IMU sensing in general, drives most of the residual error.
  • Model comparison in HAR should report the IFC or common-ground share alongside accuracy and F1, because two models with the same accuracy can fail on very different windows.
  • Future HAR data collections should adopt the proposed guidelines, especially mutually exclusive label tracks, fine-grained transition labeling, and synchronized reference video for auditability.

Reading between the lines

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

  • Editorial inference: because IFC is defined relative to a fixed panel of window-based discriminative models, the reported percentages are lower-bound estimates of ambiguity; a panel with different inductive biases could shrink them, so the numbers are evidence of dataset problems rather than exact measurements of them.
  • Editorial inference: the trinary mask could be turned into a training regularizer that downweights or ignores major segments and treats minor segments as uncertain, which would probably raise clean-window accuracy; the paper suggests this use but does not test it.
  • Editorial inference: the intersection-of-false-classifications audit is a general data-quality method that transfers to other sensor benchmarks and to any labeled time-series dataset where leave-out cross-validation is feasible.
  • Editorial inference: if the mask becomes common practice, benchmark results with and without IFC filtering will no longer be directly comparable, so dataset versions and mask versions will need to be tracked in publications.
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

4 major / 5 minor

Summary. The paper inspects six popular HAR benchmark datasets (PAMAP2, Opportunity locomotion and gesture tracks, MM-Fit, MHealth, MotionSense, WISDM) by training six model architectures (CNN, GRU, LSTM, ConvLSTM, TinyHAR, and a CPC-pretrained encoder) under Group-K-Fold cross-validation, repeated four times. It defines the Intersect of False Classifications (IFC) as the set of windows that none of the six models classifies correctly, reports the IFC percentage for each dataset, analyzes confused classes and false-classification durations, visually attributes IFC segments to three causes (ambiguous annotations, recording irregularities, and transition periods), and proposes a trinary clean/minor/major mask for filtering problematic segments. The stated goal is to show that benchmark accuracies are partly capped by dataset-inherent ambiguities rather than by model capacity alone.

Significance. If the central inference were valid, the paper would offer a practically useful quantification of annotation ambiguity across widely used HAR benchmarks and a concrete mask artifact for dataset patching. The work has genuine strengths: it is systematic (six models, four repeated training runs, grouped person/session cross-validation), it provides detailed model and preprocessing descriptions, it documents a large amount of visual inspection in the appendix, and it explicitly acknowledges its main limitation in Section 7.2. However, the significance of the quantitative IFC numbers and of the trinary mask depends on two load-bearing points that are not currently established: the correctness of the IFC calculation in Eq. (1), and the representativeness of the six-model family for the claim that some windows are inherently unclassifiable. As written, the headline claim that some segments 'cannot be correctly classified by any existing ML models' goes beyond the evidence.

major comments (4)
  1. [Section 3.3, Eq. (1), Table 4] The IFC calculation appears to overcount the intersection of false classifications. Table 4 reports 'Common Ground' and six 'Single Contributions' (percentages of windows correctly classified by only one model), and Eq. (1) computes IFC = 100% - Common Ground - sum(Single Contributions). If 'Common Ground' means windows correctly classified by all six models, then windows correctly classified by two, three, four, or five models are neither in Common Ground nor in any Single Contribution, and Eq. (1) incorrectly assigns them to the IFC. The percentages sum to 100 only if no such multi-model-correct windows exist, which is unlikely. The authors should report the union of correct classifications, or the full Venn counts, and recompute the IFC values; as they stand, the reported IFC percentages (e.g., 11.74% for PAMAP2) are upper bounds rather than the claimed intersection.
  2. [Section 3.3 and Section 7.2] The inference from 'none of the six applied models can classify a window' to 'the cause lies in the dataset itself' is an unvalidated assumption, as the authors themselves concede in Section 7.2 ('we couldn't establish a clear measure of our set of selected machine learning models sufficiently representing an appropriate intersection'). All six evaluated methods are window-based discriminative deep networks trained with Adam and cross-entropy on the same 200-sample window with 100-sample stride and the same normalization; they share strong inductive biases. The abstract's and conclusion's phrasing that such segments 'cannot be correctly classified by any existing ML models' is not supported by the experiments. The authors should either test a more architecturally diverse set (e.g., classical feature-based SVM, dilated TCN, different window sizes and strides, or a different training protocol) or reframe the IFC as 'windows that none of the six evaluated models could classify'.
  3. [Section 6, Table 6] The trinary mask is a central contribution, but the major/minor categorization is based on an ad hoc and unvalidated rule: a window is labeled 'major' if the largest gap in the sorted predicted probabilities lies between the first two probabilities, and 'minor' otherwise. The authors state in Section 6 that the automated categorization was only 'plausibly checked' through visual inspection. Since Table 6 and the proposed downstream use of the mask depend on this split, the rule should be validated quantitatively, for example by comparing mask labels with independent human re-annotation, with label noise estimates, or with the behavior of held-out models. Without such validation, the clean/minor/major percentages cannot be interpreted as measuring dataset ambiguity.
  4. [Section 4.1 and Section 6] The probability-based analysis is described only at a high level: Section 4.1 says the authors compute the 'mean calculation of false classified probability distributions across the trained models' and then select the confused class by maximum probability. It is not specified how probability distributions are aligned across models with different class orders or different hyperparameter runs, nor how the IFC windows are combined with the probability gap rule in Section 6. This makes the major/minor mask difficult to reproduce and should be clarified with precise formulas and aggregation steps.
minor comments (5)
  1. [Section 4.1] The phrase 'maximum portability selection' should be 'maximum probability selection.'
  2. [Section 5] The opening sentence contains a duplicated article: 'we only stressed the the ambiguities' should be 'we only stressed the ambiguities.'
  3. [Section 4.2] The text refers to 'MHealth and MotionSense in Fig. 10d and Fig. 10e,' but in Fig. 10 the subfigures are labeled (d) MM-FIT, (e) MHEALTH, and (f) MotionSense; the references should be adjusted accordingly.
  4. [Section 3.3] The sentence 'the origin of the wrong classification has to lay in the dataset itself' should use 'lie' instead of 'lay.'
  5. [Throughout] Dataset names are used inconsistently: 'MM-Fit' vs. 'MM-FIT' and 'MHealth' vs. 'MHEALTH' appear in different parts of the paper; please standardize the notation.

Circularity Check

2 steps flagged · score 4.0 of 10

The IFC metric is defined by six models' common failures and is then read as dataset-inherent ambiguity; the trinary mask re-encodes the same prediction probabilities as dataset labels.

  1. self definitional [Section 3 and Section 3.3, Eq. (1); echoed in Section 8]
    "we assume that the origin of the wrong classification has to lay in the dataset itself if the whole set of applied machine learning models cannot classify it properly. ... The Intersect of False Classifications (IFC) therefore represents the parts of the dataset, where none of the models could classify the window correctly. If one is able to classify the sensor signal properly, it is an indicator that such a section of the dataset contains sufficient quality to be classified."

    IFC is built by intersecting the false-classification sets of six models (Eq. 1: IFC = 100% - Common Ground - Sum of Single Contributions). The paper then interprets IFC windows as 'ambiguous annotations', 'recording execution irregularities', and 'misaligned transition periods' - dataset-inherent defects. That attribution is the Section 3 assumption ('origin ... has to lay in the dataset itself') restated as a finding, not derived from any independent label. Section 8 escalates 'six chosen state-of-the-art ML methods' to 'any existing ML models'.

  2. renaming known result [Section 6 'Proposed Solution', trinary mask definitions and Table 6]
    "We defined a trinary mask to mark each section of the dataset ... (0) clean, issue free data, confidently classified correctly across our experiments (1) minor, false classifications due to model uncertainty in experiments (2) major, strong confidence towards false classifications throughout our experiments ... This proves our assumption of uncertain and short false predicted sections due to data obscurities as opposed to longer, confident false predictions due to critical issues affecting proper classification."

    The mask's three categories are calculated from the same prediction probabilities that define IFC: clean = windows classified correctly, major = confidently wrong windows, minor = uncertain wrong windows. Labeling these bins as an 'additional dataset annotation layer' and a 'patch' renames the models' own outputs as dataset properties. The manual visual check used to validate the mask is the same inspection that generated the categories, so it cannot independently confirm that confident false predictions correspond to 'critical issues' in the data rather than to shared inductive biases of the six models.

full rationale

The paper does not rely on a load-bearing self-citation chain, a uniqueness theorem, or a fitted parameter masquerading as a prediction; its raw-signal plots and visual inspection of transitions, sensor artifacts, and activity similarity provide genuine independent evidence beyond model accuracies. However, the central move from 'none of our six models classifies this window' to 'this window is inherently ambiguous data' is an explicit assumption in Section 3 that is later reported as a finding in the abstract and conclusion, including the stronger phrasing 'cannot be correctly classified by any existing ML models'. The IFC quantity is therefore partially self-referential: it is defined from model failures and then used as evidence of dataset-inherent defects. The trinary mask similarly re-encodes the same prediction probabilities and confidence gaps as dataset-quality labels. The authors honestly flag the representativeness problem in Section 7.2, which keeps the circularity partial rather than total, and the visual signal analysis supplies independent qualitative content, so a moderate score is appropriate.

Assumptions & free parameters 6 free parameters · 5 assumptions · 2 invented entities

The central claim depends on the unverified representativeness of the six-model pool, on the choice of windowing, and on visual inspection of sensor signals as causal evidence. The IFC and the trinary mask are defined by the model outputs themselves, so their interpretation as dataset-quality measures requires the representativeness axiom. No external benchmark, code, or released mask provides independent grounding.

free parameters (6)
  • Sliding window length = 200 data points
    Chosen following common practice (refs [63,74]); all IFC and mask results depend on this windowing, and Section 7.2 acknowledges window splitting may affect classification performance.
  • Sliding window stride = 100 data points
    Chosen for comparability across datasets; affects which windows fall into the IFC.
  • Hyperparameter selection (learning rate and batch size) = Best of 9 combinations (lr 0.1/0.01/0.001, batch 64/256/1024) per dataset and model
    The best setup is selected by final validation accuracy, which is standard model selection but still a data-dependent choice that shapes the error sets feeding the IFC.
  • Major/minor gap rule = Largest probability gap between top-1 and top-2 class probabilities
    Ad hoc rule to split IFC windows into minor (uncertain) vs major (confidently wrong); no quantitative validation is provided, only informal visual inspection.
  • Maximum number of cross-validation folds = 10 groups (by merging sessions or persons)
    Chosen to limit compute; changes training set composition and therefore which windows are hard.
  • Early stopping patience = 40 epochs
    Chosen to limit resources; determines the model snapshot used for IFC extraction.
assumptions (5)
  • ad hoc to paper If none of the six selected models can classify a window, the cause lies in the dataset, not in the model family.
    Section 3.3: 'we assume that the origin of the wrong classification has to lay in the dataset itself if the whole set of applied machine learning models cannot classify it properly.' This is load-bearing and not established; Section 7.2 admits the model set's representativeness is unmeasured.
  • ad hoc to paper The chosen six model architectures sufficiently span the space of reasonable HAR classifiers.
    Section 3.2 selects 'milestone' architectures, but no argument or measure shows they are representative; acknowledged in Section 7.2.
  • domain assumption Raw sensor-signal inspection can identify causes of misclassification without synchronized video ground truth.
    Section 5 derives causal categories (transitions, data quality, activity uncertainty) from visual inspection of sensor plots; Section 7.2 notes video is unavailable for most datasets.
  • domain assumption A 200-sample sliding window is a faithful unit for labeling activities.
    Section 3.1 uses one fixed window size; Section 7.2 acknowledges window splitting may be unfortunate for some activity patterns.
  • standard math The set-theoretic partition of windows into common ground, single contributions, and IFC is well-defined.
    Eq. (1) relies on a partition, but the definition of 'Common Ground' is ambiguous and Table 4 is internally inconsistent for Oppo-Gest.
invented entities (2)
  • Intersect of False Classifications (IFC)
    purpose: Metric for windows that no evaluated model classifies correctly; used as a dataset ambiguity measure
    Defined entirely from the six models' errors; no independent ground truth or external benchmark validates that IFC equals inherent ambiguity.
  • Trinary IFC mask (0 clean, 1 minor, 2 major)
    purpose: Dataset patching layer to filter or reweight IFC sections during training or evaluation
    Labels are generated from the same model probability outputs that define the IFC; the categorization rule is heuristic and the mask has not been released for independent use.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Beyond Confusion: A Fine-grained Dialectical Examination of Human Activity Recognition Benchmark Datasets." pith.science (2026). https://pith.science/paper/AHAHGUMC

@misc{pith2026241209037,
  author       = {Pith},
  title        = {Pith review of: Beyond Confusion: A Fine-grained Dialectical Examination of Human Activity Recognition Benchmark Datasets},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AHAHGUMC}},
  note         = {Machine review of arXiv:2412.09037}
}
read the original abstract

The research of machine learning (ML) algorithms for human activity recognition (HAR) has made significant progress with publicly available datasets. However, most research prioritizes statistical metrics over examining negative sample details. While recent models like transformers have been applied to HAR datasets with limited success from the benchmark metrics, their counterparts have effectively solved problems on similar levels with near 100% accuracy. This raises questions about the limitations of current approaches. This paper aims to address these open questions by conducting a fine-grained inspection of six popular HAR benchmark datasets. We identified for some parts of the data, none of the six chosen state-of-the-art ML methods can correctly classify, denoted as the intersect of false classifications (IFC). Analysis of the IFC reveals several underlying problems, including ambiguous annotations, irregularities during recording execution, and misaligned transition periods. We contribute to the field by quantifying and characterizing annotated data ambiguities, providing a trinary categorization mask for dataset patching, and stressing potential improvements for future data collections.

Figures

Figures reproduced from arXiv: 2412.09037 by the authors.

Figure 1
Figure 1. We aim to examine HAR research beyond the typical approach focusing on the statistical metrics and confusion [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Extracting the Overlap of False Classification across the models, then merging the overlapping slinging windows into [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Chord diagrams of PAMAP2, with null class added on the left and null class removed on the right for clarity [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗
Figures from the paper (17 more)
Figure 4
Figure 4. Figure 4: Chord diagrams and table of class distribution and confusion in Opportunity with locomotion labels. [PITH_FULL_IMAGE:figures/full_fig_p012_4.png]
Figure 5
Figure 5. Figure 5: Chord diagrams and table of class distribution and confusion in Opportunity with gesture labels. [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: Chord diagrams and table of class distribution and confusion in MM-FIT. [PITH_FULL_IMAGE:figures/full_fig_p013_6.png]
Figure 7
Figure 7. Figure 7: Chord diagrams and table of class distribution and confusion in MHealth. [PITH_FULL_IMAGE:figures/full_fig_p014_7.png]
Figure 8
Figure 8. Figure 8: Chord diagrams and table of class distribution and confusion in MotionSense. [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]
Figure 9
Figure 9. Figure 9: Chord diagrams and table of class distribution and confusion in WISDM. [PITH_FULL_IMAGE:figures/full_fig_p015_9.png]
Figure 10
Figure 10. Figure 10: Histograms (log-scale) for each dataset, representing the occurrence of continuous false classified windows. [PITH_FULL_IMAGE:figures/full_fig_p016_10.png]
Figure 11
Figure 11. Figure 11: Exemplary view (PAMAP2 of Participant 1) visualizing the acceleration data of the three IMUs underlaid with correct [PITH_FULL_IMAGE:figures/full_fig_p017_11.png]
Figure 12
Figure 12. Figure 12: Confusion between Standing and Ironing in PAMAP2; the IFC shows false classified activities during transition the false classified classes, there is a clear change in acceleration data due to movements. Even though the long section obtains an almost consistent signal,…
Figure 13
Figure 13. Figure 13: Periodic noise resulting in false classification in MHealth dataset ( [PITH_FULL_IMAGE:figures/full_fig_p019_13.png]
Figure 14
Figure 14. Figure 14: Ironing Activity of PAMAP2 compared between participants 8 (left) and 7 (right). The Magnetometer of Participant 7 [PITH_FULL_IMAGE:figures/full_fig_p020_14.png]
Figure 15
Figure 15. Figure 15: Sensor Data and False Classifications of PAMAP2 [PITH_FULL_IMAGE:figures/full_fig_p028_15.png]
Figure 16
Figure 16. Figure 16: Sensor Data and False Classifications of OPPORTUNITY [PITH_FULL_IMAGE:figures/full_fig_p029_16.png]
Figure 17
Figure 17. Figure 17: Sensor Data and False Classifications of MM-FIT [PITH_FULL_IMAGE:figures/full_fig_p030_17.png]
Figure 18
Figure 18. Figure 18: Sensor Data and False Classifications of MHEALTH [PITH_FULL_IMAGE:figures/full_fig_p031_18.png]
Figure 19
Figure 19. Figure 19: Sensor Data and False Classifications of MotionSense [PITH_FULL_IMAGE:figures/full_fig_p032_19.png]
Figure 20
Figure 20. Figure 20: Sensor Data and False Classifications of WISDM [PITH_FULL_IMAGE:figures/full_fig_p033_20.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. TinierHAR: Towards Ultra-Lightweight Deep Learning Models for Efficient Human Activity Recognition on Edge Devices

    cs.CV 2025-07 conditional novelty 5.0 of 10

    TinierHAR is an ultra-lightweight HAR model that matches TinyHAR's F1 score with 2.7x fewer parameters and 6.4x fewer MACs across 14 datasets.

  2. FedFitTech: A Baseline in Federated Learning for Fitness Tracking

    cs.LG 2025-06 conditional novelty 4.0 of 10

    An open-source Flower-based federated learning baseline for fitness tracking, plus a case study showing client-side early stopping cuts communication 13% with a 1% F1 drop.

Reference graph

Works this paper leans on

96 extracted references · 60 canonical work pages · cited by 2 Pith papers

  1. [1]

    Alireza Abedin, Mahsa Ehsanpour, Qinfeng Shi, Hamid Rezatofighi, and Damith C Ranasinghe. 2021. Attend and discriminate: Beyond the state-of-the-art for human activity recognition using wearable sensors. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 5, 1 (2021), 1–22

    2021

  2. [2]

    Mohammed AA Al-Qaness, Abdelghani Dahou, Mohamed Abd Elaziz, and AM Helmi. 2022. Multi-ResAtt: Multilevel residual network with attention for human activity recognition using wearable sensors. IEEE Transactions on Industrial Informatics 19, 1 (2022), 144–152

    2022

  3. [3]

    Mohammad Abu Alsheikh, Ahmed Selim, Dusit Niyato, Linda Doyle, Shaowei Lin, and Hwee Pink Tan. 2016. Deep activity recognition models with triaxial accelerometers. In 30th AAAI Conference on Artificial Intelligence, AAAI 2016 . AI Access Foundation, 8–13

    2016

  4. [4]

    Davide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra, and Jorge L Reyes-Ortiz. 2012. Human activity recognition on smartphones using a multiclass hardware-friendly support vector machine. In International workshop on ambient assisted living . Springer, 216–223

    2012

  5. [5]

    Davide Anguita, Alessandro Ghio, Luca Oneto, Xavier Parra, Jorge Luis Reyes-Ortiz, et al. 2013. A public domain dataset for human activity recognition using smartphones.. In Esann, Vol. 3. 3

    2013

  6. [6]

    Ferhat Attal, Samer Mohammed, Mariam Dedabrishvili, Faicel Chamroukhi, Latifa Oukhellou, and Yacine Amirat. 2015. Physical human activity recognition using wearable sensors. Sensors 15, 12 (2015), 31314–31338

    2015

  7. [7]

    Lei Bai, Lina Yao, Xianzhi Wang, Salil S Kanhere, Bin Guo, and Zhiwen Yu. 2020. Adversarial multi-view networks for activity recognition. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 4, 2 (2020), 1–22. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol., Vol. 0, No. 0, Article 0. Publication date: 2024. Beyond Conf...

    2020

  8. [8]

    Oresti Baños, Miguel Damas, Héctor Pomares, Ignacio Rojas, Máté Attila Tóth, and Oliver Amft. 2012. A benchmark dataset to evaluate sensor displacement in activity recognition. In Proceedings of the 2012 ACM Conference on Ubiquitous Computing . 1026–1035

    2012

Show all 96 references
  1. [9]

    Oresti Banos, Rafael Garcia, Juan A Holgado-Terriza, Miguel Damas, Hector Pomares, Ignacio Rojas, Alejandro Saez, and Claudia Villalonga. 2014. mHealthDroid: a novel framework for agile development of mobile health applications. In Ambient Assisted Living and Daily Activities:...

    2014

  2. [10]

    Sizhen Bian, Mengxi Liu, Bo Zhou, and Paul Lukowicz. 2022. The state-of-the-art sensing techniques in human activity recognition: A survey. Sensors 22, 12 (2022), 4596

    2022

  3. [11]

    Sizhen Bian and Michele Magno. 2023. Evaluating Spiking Neural Network on Neuromorphic Platform For Human Activity Recognition. In Proceedings of the 2023 ACM International Symposium on Wearable Computers . 82–86

    2023

  4. [12]

    Andreas Bulling, Ulf Blanke, and Bernt Schiele. 2014. A tutorial on human activity recognition using body-worn inertial sensors. ACM Computing Surveys (CSUR) 46, 3 (2014), 1–33

    2014

  5. [13]

    Rich Caruana. 1997. Multitask Learning. Machine Learning 28, 1 (01 Jul 1997), 41–75. https://doi.org/10.1023/A:1007379606734

    1997 doi

  6. [14]

    S Chan Chang, R Walmsley, J Gershuny, T Harms, E Thomas, K Milton, P Kelly, C Foster, A Wong, N Gray, et al. 2021. Capture-24: Activity tracker dataset for human activity recognition. (2021)

    2021

  7. [15]

    Ricardo Chavarriaga, Hesam Sagha, Alberto Calatroni, Sundara Tejaswi Digumarti, Gerhard Tröster, José del R Millán, and Daniel Roggen. 2013. The Opportunity challenge: A benchmark database for on-body sensor-based activity recognition. Pattern Recognition Letters 34, 15 (2013)...

    2013

  8. [16]

    Chen Chen, Roozbeh Jafari, and Nasser Kehtarnavaz. 2015. UTD-MHAD: A multimodal dataset for human action recognition utilizing a depth camera and a wearable inertial sensor. In 2015 IEEE International conference on image processing (ICIP) . IEEE, 168–172

    2015

  9. [17]

    Ling Chen, Yi Zhang, and Liangying Peng. 2020. Metier: A deep multi-task learning based activity and user recognition model using wearable sensors. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 4, 1 (2020), 1–18

    2020

  10. [18]

    Dongzhou Cheng, Lei Zhang, Can Bu, Hao Wu, and Aiguo Song. 2023. Learning hierarchical time series data augmentation invariances via contrastive supervision for human activity recognition. Knowledge-Based Systems 276 (2023), 110789

    2023

  11. [19]

    Shohreh Deldari, Hao Xue, Aaqib Saeed, Daniel V Smith, and Flora D Salim. 2022. Cocoa: Cross modality contrastive learning for sensor data. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 6, 3 (2022), 1–28

    2022

  12. [20]

    Jia Deng, Wei Dong, Richard Socher, Li-Jia Li, Kai Li, and Li Fei-Fei. 2009. Imagenet: A large-scale hierarchical image database. In 2009 IEEE conference on computer vision and pattern recognition . Ieee, 248–255

    2009

  13. [21]

    Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Sylvain Gelly, Jakob Uszkoreit, and Neil Houlsby. 2021. An Image is Worth 16x16 Words: Transformers for Image Recognit...

    2021

  14. [22]

    Walter Durka and Stefan G Michalski. 2012. Daphne: a dated phylogeny of a large European flora for phylogenetically informed ecological analyses: Ecological Archives E093-214. Ecology 93, 10 (2012), 2297–2297

    2012

  15. [23]

    Kilian Forster, Daniel Roggen, and Gerhard Troster. 2009. Unsupervised classifier self-calibration through repeated context occurences: Is there robustness against sensor displacement to gain?. In 2009 international symposium on wearable computers . IEEE, 77–84

    2009

  16. [24]

    Vitor Fortes Rey, Kamalveer Kaur Garewal, and Paul Lukowicz. 2021. Translating videos into synthetic training data for wearable sensor-based activity recognition systems using residual deep convolutional networks. Applied Sciences 11, 7 (2021), 3094

    2021

  17. [25]

    Vitor Fortes Rey, Dominique Nshimyimana, and Paul Lukowicz. 2023. Don’t freeze: Finetune encoders for better Self-Supervised HAR. In Adjunct Proceedings of the 2023 ACM International Joint Conference on Pervasive and Ubiquitous Computing & the 2023 ACM International Symposium ...

    2023

  18. [26]

    Vitor Fortes Rey, Sungho Suh, and Paul Lukowicz. 2022. Learning from the Best: Contrastive Representations Learning Across Sensor Locations for Wearable Activity Recognition. In Proceedings of the 2022 ACM International Symposium on Wearable Computers . 28–32

    2022

  19. [27]

    Wenbin Gao, Lei Zhang, Qi Teng, Jun He, and Hao Wu. 2021. DanHAR: Dual attention network for multimodal human activity recognition using wearable sensors. Applied Soft Computing 111 (2021), 107728

    2021

  20. [28]

    Daniel Geißler, Esther Friederike Zahn, Hymalai Bello, Lala Shakti Swarup Ray, Emil Woop, Bo Zhou, Paul Lukowicz, and Gesche Joost. 2023. MoCa’Collection: Normalizing Dynamic Textile Geometry with Capacitive Sensing in Design Centric Wearables. In Adjunct Proceedings of the 20...

    2023

  21. [29]

    Daniel Geißler, Bo Zhou, Hymalai Bello, Joanna Sorysz, Lala Ray, Hamraz Javaheri, Matthias Rüb, Jan Herbst, Esther Zahn, Emil Woop, et al. 2024. Embedding textile capacitive sensing into smart wearables as a versatile solution for human motion capturing. Scientific Reports 14,...

    2024

  22. [30]

    Daniel Geißler, Bo Zhou, Mengxi Liu, Sungho Suh, and Paul Lukowicz. 2024. The Power of Training: How Different Neural Network Setups Influence the Energy Demand. In International Conference on Architecture of Computing Systems . Springer, 33–47

    2024

  23. [31]

    Daniel Geißler, Bo Zhou, Paul Lukowicz, and RPTU Kaiserslautern-Landau. 2023. Latent Inspector: An Interactive Tool for Probing Neural Network Behaviors Through Arbitrary Latent Activation.. In IJCAI. 7127–7130. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol., Vol. 0, No...

    2023

  24. [32]

    Jonathan Gershuny, Teresa Harms, Aiden Doherty, Emma Thomas, Karen Milton, Paul Kelly, and Charlie Foster. 2020. Testing Self-Report Time-Use Diaries against Objective Instruments in Real Time. Sociological Methodology 50, 1 (2020), 318–349. https: //doi.org/10.1177/0081175019...

    2020 doi

  25. [33]

    Hristijan Gjoreski, Mathias Ciliberto, Lin Wang, Francisco Javier Ordonez Morales, Sami Mekki, Stefan Valentin, and Daniel Roggen

  26. [34]

    Yu Guan and Thomas Plötz. 2017. Ensembles of deep lstm learners for activity recognition using wearables. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 1, 2 (2017), 1–28

    2017

  27. [35]

    Nils Y Hammerla, Reuben Kirkham, Peter Andras, and Thomas Ploetz. 2013. On preserving statistical characteristics of accelerometry data using their empirical cumulative distribution. In Proceedings of the 2013 international symposium on wearable computers . 65–68

    2013

  28. [36]

    Harish Haresamudram, Irfan Essa, and Thomas Plötz. 2021. Contrastive Predictive Coding for Human Activity Recognition. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol. 5, 2, Article 65 (jun 2021), 26 pages. https://doi.org/10.1145/3463506

    2021 doi

  29. [37]

    Harish Haresamudram, Irfan Essa, and Thomas Plötz. 2022. Assessing the state of self-supervised human activity recognition using wearables. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 6, 3 (2022), 1–47

    2022

  30. [38]

    Harish Haresamudram, Irfan Essa, and Thomas Plötz. 2022. Assessing the State of Self-Supervised Human Activity Recognition using Wearables. arXiv:2202.12938 [eess.SP]

    2022 arXiv

  31. [39]

    Harish Haresamudram, Irfan Essa, and Thomas Plötz. 2023. Investigating Enhancements to Contrastive Predictive Coding for Human Activity Recognition. In 2023 IEEE International Conference on Pervasive Computing and Communications (PerCom) . 232–241. https: //doi.org/10.1109/PER...

    2023

  32. [40]

    Ramin Hasani, Mathias Lechner, Alexander Amini, Lucas Liebenwein, Aaron Ray, Max Tschaikowski, Gerald Teschl, and Daniela Rus

  33. [41]

    Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long short-term memory. Neural computation 9, 8 (1997), 1735–1780

    1997

  34. [42]

    Alexander Hoelzemann, Marius Bock, and Kristof Van Laerhoven. 2024. Evaluation of Video-Assisted Annotation of Human IMU Data Across Expertise, Datasets, and Tools. In 2024 IEEE International Conference on Pervasive Computing and Communications Workshops and other Affiliated E...

    2024

  35. [43]

    Tâm Huynh, Mario Fritz, and Bernt Schiele. 2008. Discovery of activity patterns using topic models. InProceedings of the 10th international conference on Ubiquitous computing . 10–19

    2008

  36. [44]

    Wenchao Jiang and Zhaozheng Yin. 2015. Human activity recognition using wearable sensors by deep convolutional neural networks. In Proceedings of the 23rd ACM international conference on Multimedia . 1307–1310

    2015

  37. [45]

    Fabian Karl and Ansgar Scherp. 2022. Transformers are short text classifiers: A study of inductive short text classifiers on benchmarks and real-world datasets. arXiv preprint arXiv:2211.16878 (2022)

    2022 arXiv

  38. [46]

    Nobuo Kawaguchi, Nobuhiro Ogawa, Yohei Iwasaki, Katsuhiko Kaji, Tsutomu Terada, Kazuya Murao, Sozo Inoue, Yoshihiro Kawahara, Yasuyuki Sumi, and Nobuhiko Nishio. 2011. HASC Challenge: gathering large scale human activity corpus for the real-world activity understandings. In Pr...

    2011

  39. [47]

    Shian-Ru Ke, Hoang Le Uyen Thuc, Yong-Jin Lee, Jenq-Neng Hwang, Jang-Hee Yoo, and Kyoung-Ho Choi. 2013. A review on video-based human activity recognition. Computers 2, 2 (2013), 88–131

    2013

  40. [48]

    Alex Krizhevsky, Geoffrey Hinton, et al. 2009. Learning multiple layers of features from tiny images. (2009)

    2009

  41. [49]

    Jennifer R Kwapisz, Gary M Weiss, and Samuel A Moore. 2011. Activity recognition using cell phone accelerometers. ACM SigKDD Explorations Newsletter 12, 2 (2011), 74–82

    2011

  42. [50]

    Hyeokhyen Kwon, Gregory D Abowd, and Thomas Plötz. 2019. Handling annotation uncertainty in human activity recognition. In Proceedings of the 2019 ACM International Symposium on Wearable Computers . 109–117

    2019

  43. [51]

    Hyeokhyen Kwon, Catherine Tong, Harish Haresamudram, Yan Gao, Gregory D Abowd, Nicholas D Lane, and Thomas Ploetz. 2020. Imutube: Automatic extraction of virtual on-body accelerometry from video for human activity recognition. Proceedings of the ACM on Interactive, Mobile, Wea...

    2020

  44. [52]

    Hyeokhyen Kwon, Bingyao Wang, Gregory D Abowd, and Thomas Plötz. 2021. Approaching the real-world: Supporting activity recognition training with virtual imu data. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 5, 3 (2021), 1–32

    2021

  45. [53]

    Paula Lago, Moe Matsuki, Kohei Adachi, and Sozo Inoue. 2021. Using additional training sensors to improve single-sensor complex activity recognition. In Proceedings of the 2021 ACM International Symposium on Wearable Computers . 18–22

    2021

  46. [54]

    Oscar D Lara and Miguel A Labrador. 2012. A survey on human activity recognition using wearable sensors. IEEE communications surveys & tutorials 15, 3 (2012), 1192–1209

    2012

  47. [55]

    Mengxi Liu, Vitor Fortes Rey, Yu Zhang, Lala Shakti Swarup Ray, Bo Zhou, and Paul Lukowicz. 2024. iMove: Exploring Bio-impedance Sensing for Fitness Activity Recognition. arXiv preprint arXiv:2402.09445 (2024). Proc. ACM Interact. Mob. Wearable Ubiquitous Technol., Vol. 0, No....

    2024 arXiv

  48. [56]

    Saif Mahmud, M Tanjid Hasan Tonmoy, Kishor Kumar Bhaumik, AKM Mahbubur Rahman, M Ashraful Amin, Mohammad Shoyaib, Muhammad Asif Hossain Khan, and Amin Ahsan Ali. 2020. Human Activity Recognition from Wearable Sensor Data Using Self-Attention. In ECAI 2020. IOS Press, 1332–1339

    2020

  49. [57]

    Mohammad Malekzadeh, Richard G Clegg, Andrea Cavallaro, and Hamed Haddadi. 2018. Protecting sensory data against sensitive inferences. In Proceedings of the 1st Workshop on Privacy by Design in Distributed Systems . 1–6

    2018

  50. [58]

    Ronald Mutegeki and Dong Seog Han. 2020. A CNN-LSTM Approach to Human Activity Recognition. In 2020 International Conference on Artificial Intelligence in Information and Communication (ICAIIC) . 362–366. https://doi.org/10.1109/ICAIIC48513.2020.9065078

    2020

  51. [59]

    Kotaro Nakano and Basabi Chakraborty. 2017. Effect of dynamic feature for human activity recognition using smartphone sensors. In 2017 IEEE 8th International Conference on A wareness Science and Technology (iCAST) . IEEE, 539–543

    2017

  52. [60]

    Duc-Anh Nguyen, Cuong Pham, and Nhien-An Le-Khac. 2023. Virtual Fusion with Contrastive Learning for Single Sensor-based Activity Recognition. arXiv preprint arXiv:2312.02185 (2023)

    2023 arXiv

  53. [61]

    Aaron van den Oord, Yazhe Li, and Oriol Vinyals. 2018. Representation learning with contrastive predictive coding. arXiv preprint arXiv:1807.03748 (2018)

    2018 arXiv

  54. [62]

    Francisco Javier Ordóñez and Daniel Roggen. 2016. Deep convolutional and lstm recurrent neural networks for multimodal wearable activity recognition. Sensors 16, 1 (2016), 115

    2016

  55. [63]

    Francisco Javier Ordóñez and Daniel Roggen. 2016. Deep Convolutional and LSTM Recurrent Neural Networks for Multimodal Wearable Activity Recognition. Sensors 16, 1 (2016). https://doi.org/10.3390/s16010115

    2016 doi

  56. [64]

    Thomas Plötz, Nils Y Hammerla, and Patrick L Olivier. 2011. Feature learning for activity recognition in ubiquitous computing. In Twenty-second international joint conference on artificial intelligence

    2011

  57. [65]

    Daniele Ravi, Charence Wong, Benny Lo, and Guang-Zhong Yang. 2016. A deep learning approach to on-node sensor data analytics for mobile or wearable devices. IEEE journal of biomedical and health informatics 21, 1 (2016), 56–64

    2016

  58. [66]

    Attila Reiss and Didier Stricker. 2012. Introducing a new benchmarked dataset for activity monitoring. In 2012 16th international symposium on wearable computers . IEEE, 108–109

    2012

  59. [67]

    Vitor Fortes Rey, Peter Hevesi, Onorina Kovalenko, and Paul Lukowicz. 2019. Let there be IMU data: generating training data for wearable, motion sensor based activity recognition from monocular RGB videos. In Adjunct proceedings of the 2019 ACM international joint conference o...

    2019

  60. [68]

    Jorge-L Reyes-Ortiz, Luca Oneto, Albert Samà, Xavier Parra, and Davide Anguita. 2016. Transition-aware human activity recognition using smartphones. Neurocomputing 171 (2016), 754–767

    2016

  61. [69]

    Taoran Sheng and Manfred Huber. 2020. Weakly supervised multi-task representation learning for human activity analysis using wearables. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 4, 2 (2020), 1–18

    2020

  62. [70]

    Lucas Smaira, João Carreira, Eric Noland, Ellen Clancy, Amy Wu, and Andrew Zisserman. 2020. A Short Note on the Kinetics-700-2020 Human Action Dataset. arXiv:2010.10864 [cs.CV]

    2020 arXiv

  63. [71]

    Siddharth Srivastava and Gaurav Sharma. 2024. Omnivec: Learning robust representations with cross modal sharing. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision . 1236–1248

    2024

  64. [72]

    David Strömbäck, Sangxia Huang, and Valentin Radu. 2020. Mm-fit: Multimodal deep learning for automatic exercise logging across sensing devices. Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies 4, 4 (2020), 1–22

    2020

  65. [73]

    Sungho Suh, Vitor Fortes Rey, and Paul Lukowicz. 2022. Adversarial deep feature extraction network for user independent human activity recognition. In 2022 IEEE International Conference on Pervasive Computing and Communications (PerCom) . IEEE, 217–226

    2022

  66. [74]

    Sungho Suh, Vitor Fortes Rey, and Paul Lukowicz. 2023. TASKED: Transformer-based Adversarial learning for human activity recognition using wearable sensors via Self-KnowledgE Distillation. Knowledge-Based Systems 260 (2023), 110143

    2023

  67. [75]

    Timo Sztyler and Heiner Stuckenschmidt. 2016. On-body localization of wearable devices: An investigation of position-aware activity recognition. In 2016 IEEE International Conference on Pervasive Computing and Communications (PerCom) . IEEE, 1–9

    2016

  68. [76]

    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

  69. [77]

    Andrés Tello, Victoria Degeler, and Alexander Lazovik. 2024. Too Good To Be True: accuracy overestimation in (re) current practices for Human Activity Recognition. In 2024 IEEE International Conference on Pervasive Computing and Communications Workshops and other Affiliated Ev...

    2024

  70. [78]

    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

  71. [79]

    George Vavoulas, Charikleia Chatzaki, Thodoris Malliotakis, Matthew Pediaditis, and Manolis Tsiknakis. 2016. The mobiact dataset: Recognition of activities of daily living using smartphones. In International conference on information and communication technologies for ageing w...

    2016

  72. [80]

    Jindong Wang, Yiqiang Chen, Shuji Hao, Xiaohui Peng, and Lisha Hu. 2019. Deep learning for sensor-based activity recognition: A survey. Pattern recognition letters 119 (2019), 3–11. Proc. ACM Interact. Mob. Wearable Ubiquitous Technol., Vol. 0, No. 0, Article 0. Publication da...

    2019

  73. [81]

    Shuoyuan Wang, Lei Zhang, Xing Wang, Wenbo Huang, Hao Wu, and Aiguo Song. 2024. PatchHAR: A MLP-Like Architecture for Efficient Activity Recognition Using Wearables. IEEE Transactions on Biometrics, Behavior, and Identity Science (2024)

    2024

  74. [82]

    Jamie A Ward, Paul Lukowicz, and Hans W Gellersen. 2011. Performance metrics for activity recognition.ACM Transactions on Intelligent Systems and Technology (TIST) 2, 1 (2011), 1–23

    2011

  75. [83]

    Christoph Wieland and Victor Pankratius. 2023. TinyGraphHAR: Enhancing human activity recognition with graph neural networks. In 2023 IEEE World AI IoT Congress (AIIoT). IEEE, 0047–0054

    2023

  76. [84]

    Matthew Willetts, Sven Hollowell, Louis Aslett, Chris Holmes, and Aiden Doherty. 2018. Statistical machine learning of sleep and physical activity phenotypes from sensor data in 96,220 UK Biobank participants. Scientific reports 8, 1 (2018), 7961

    2018

  77. [85]

    Fanyi Xiao, Ling Pei, Lei Chu, Danping Zou, Wenxian Yu, Yifan Zhu, and Tao Li. 2021. A deep learning method for complex human activity recognition using virtual wearable sensors. In Spatial Data and Intelligence: First International Conference, SpatialDI 2020, Virtual Event, M...

    2021

  78. [86]

    Huiyuan Yang, Han Yu, Kusha Sridhar, Thomas Vaessen, Inez Myin-Germeys, and Akane Sano. 2022. More to Less (M2L): Enhanced Health Recognition in the Wild with Reduced Modality of Wearable Sensors. In 2022 44th Annual International Conference of the IEEE Engineering in Medicine...

    2022

  79. [87]

    Jianbo Yang, Minh Nhut Nguyen, Phyo Phyo San, Xiaoli Li, and Shonali Krishnaswamy. 2015. Deep convolutional neural networks on multichannel time series for human activity recognition.. In Ijcai, Vol. 15. Buenos Aires, Argentina, 3995–4001

    2015

  80. [88]

    Jianbo Yang, Minh Nhut Nguyen, Phyo Phyo San, Xiao Li Li, and Shonali Krishnaswamy. 2015. Deep convolutional neural networks on multichannel time series for human activity recognition. In Twenty-fourth international joint conference on artificial intelligence . 3995–4001

    2015

  81. [89]

    Hang Yuan, Shing Chan, Andrew P Creagh, Catherine Tong, Aidan Acquah, David A Clifton, and Aiden Doherty. 2024. Self-supervised learning for human activity recognition using 700,000 person-days of wearable data. NPJ Digital Medicine 7, 1 (2024), 91

    2024

  82. [90]

    Piero Zappi, Clemens Lombriser, Thomas Stiefmeier, Elisabetta Farella, Daniel Roggen, Luca Benini, and Gerhard Tröster. 2008. Activity recognition from on-body sensors: accuracy-power trade-off by dynamic sensor selection. In Wireless Sensor Networks: 5th European Conference, ...

    2008

  83. [91]

    Ming Zeng, Haoxiang Gao, Tong Yu, Ole J Mengshoel, Helge Langseth, Ian Lane, and Xiaobing Liu. 2018. Understanding and improving recurrent networks for human activity recognition by continuous attention. In Proceedings of the 2018 ACM international symposium on wearable comput...

    2018

  84. [92]

    Licheng Zhang, Xihong Wu, and Dingsheng Luo. 2015. Recognizing human activities from raw accelerometer data using deep neural networks. In 2015 IEEE 14th International conference on machine learning and applications (ICMLA) . IEEE, 865–870

    2015

  85. [93]

    Bo Zhou, Daniel Geissler, Marc Faulhaber, Clara Elisabeth Gleiss, Esther Friederike Zahn, Lala Shakti Swarup Ray, David Gamarra, Vitor Fortes Rey, Sungho Suh, Sizhen Bian, et al . 2023. Mocapose: Motion capturing with textile-integrated capacitive sensors in loose-fitting smar...

    2023

  86. [94]

    Yexu Zhou, Haibin Zhao, Yiran Huang, Till Riedel, Michael Hefenbrock, and Michael Beigl. 2022. TinyHAR: A lightweight deep learning model designed for human activity recognition. In Proceedings of the 2022 ACM International Symposium on Wearable Computers . 89–93. Proc. ACM In...

    2022

  87. [2018]

    IEEE Access 6 (2018), 42592–42604

    The university of sussex-huawei locomotion and transportation dataset for multimodal analytics with mobile devices. IEEE Access 6 (2018), 42592–42604

    2018

  88. [2022]

    Nature Machine Intelligence 4, 11 (2022), 992–1003

    Closed-form continuous-time neural networks. Nature Machine Intelligence 4, 11 (2022), 992–1003

    2022

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.