Pith. sign in

REVIEW 4 major objections 6 minor 43 references

XMTC: Explainable Early Classification of Multivariate Time Series in Reach-to-Grasp Hand Kinematics

T0 review · 4 major / 6 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read XMTC claims that a reach-to-grasp target can be predicted from hand kinematics before the hand touches the object, with an interval-forest classifier reaching 73% accuracy at a 360-step window and a visualization tool exposing why…

desk verdict A useful visual analytics tool, but the before-grasp accuracy claim needs a user-independent evaluation before it can be believed. read the letter →

arxiv 2502.04398 v1 pith:CJBB4UVF submitted 2025-02-06 cs.LG cs.GRcs.HC

Reyhaneh Sabbagh Gol , Dimitar Valkov , Lars Linsen This is my paper · ORCID
classification cs.LGcs.GRcs.HC
keywords earlytimeseriesclassificationmultivariatereach-to-grasphandkinematicsDrCIFexplainableAIvisualanalyticsobjectprediction
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 asks whether a person's intended grasp target can be read from hand motion early enough to be useful, and it answers yes for controlled reach-to-grasp recordings. In the strongest scenario, the classifier reaches 73.02% accuracy at time step 360, before any tested participant has touched the object, and 91.01% at time step 640. The same analysis shows when classes become separable, which objects remain confused, and which finger-aperture features carry the signal. The authors wrap the classifier in XMTC, a visual tool that connects accuracy-over-window plots, confusion heatmaps, per-series prediction heatmaps, and partial dependence plots, so a designer can choose an earliness/accuracy trade-off with evidence in view.

What carries the argument

The load-bearing classifier is DrCIF (Diverse Representation Canonical Interval Forest), an ensemble of 200 decision trees that draw random intervals from three representations of the series, namely the original series, first-order differences, and periodograms, and compute 29 statistical features including the catch22 set over each interval, with random dimension and feature subset selection per tree and majority voting at the end. Around this classifier, the paper builds a moving-window protocol: one model per prefix $[0,10], [0,20], \ldots$ of the time series, so the accuracy curve itself becomes the earliness/accuracy axis. XMTC then maps each analysis task to a coordinated view, using the accuracy plot with series-length histograms, confusion-matrix heatmaps, a per-series class-probability heatmap, and partial dependence plots, so global accuracy, confusion evolution, individual trajectories, and feature impact are inspectable from the same trained models.

What would settle it

Run the same windowed DrCIF training on Experiment 3 with a strict user-exclusive split, holding out all trials of a set of participants rather than random trials, and read the accuracy at window 360 and window 640; if before-touch accuracy falls to chance level or well below the leave-one-out mean, the central early-prediction claim does not transfer to unseen users.

Watch

Extended reading notes

Core claim

On its own terms, the paper establishes that early object prediction from reach-to-grasp kinematics is not only possible but diagnosable, where the kinematic input is a 12-dimensional time series of aperture vectors: the x, y, and z components of the thumb-to-index, thumb-to-middle, thumb-to-ring, and thumb-to-little-finger distances. The intended object among eight condition classes is predicted from the first $i\cdot 10$ time steps of this time series, with short series stretched to a common window length for training. Experiment 3 improves from 29.97% at window 10 to 73.02% at window 360, the last step before any object is grasped, and to 91.01% by window 640, stabilizing near 93.19% at the full length. Experiment 1 reaches about 71% at the recommended window 380 and 85.4% at the full 1436-step window, while Experiment 2 stalls near 60%, which the authors attribute to participants not attending to object identity. The explanations identify persistent confusions between similar objects and single out, for example, the x-components of thumb-middle and thumb-ring apertures as the features that lift one class from 74% to 100% accuracy.

Load-bearing premise

The load-bearing premise is that accuracy measured on a random 20% holdout of trials predicts how the model will do on a person whose movement style was never in training; the paper's own leave-one-out test shows that switching to held-out users drops the recommended Experiment 1 model from 71.09% to an average of 60.62% with a standard deviation near 10 percentage points.

Editorial extensions

If this is right

  • In reach-to-grasp interfaces, the intended object can be named before the hand touches it, at 73% accuracy under the paper's evaluation in Experiment 3.
  • A tool user can choose the operating point explicitly: window 360 for maximal earliness, or window 640 for the steady state near 91% accuracy.
  • Persistent confusions are structural, not random: similar objects such as bottle and cup or knife and pen are the last to separate, so a system can treat those pairs as expected ambiguities.
  • Feature-impact analysis points to specific aperture components, such as the x-components of thumb-middle and thumb-ring apertures, as the discriminative signal, which can guide sensor choice.
  • Because the explanation views are global, post-hoc, and model-agnostic, the same visual analysis can be attached to a different classifier without changing the tool's logic.

Reading between the lines

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

  • The authors leave implicit that deployment accuracy for a new user is the user-independent number: their own leave-one-out test drops the recommended Experiment 1 model from 71.09% to a mean of 60.62%, so an actual interface should be evaluated by held-out users, not random trials.
  • At the 960 Hz motion-capture rate, the 360-step window is 0.375 seconds into the reach; if the before-touch accuracy transfers, a system could begin acting about a third of a second before contact, which is enough to mask typical interaction latency.
  • A direct extension is to ablate the x-components of thumb-middle and thumb-ring apertures: the partial dependence analysis implies they carry the decisive early signal for at least one class, so a reduced sensor setup might preserve much of the accuracy.
  • The tool's accuracy-over-window plot is most informative when series lengths vary; in fixed-length multivariate time series, the early axis would need a different operationalization, such as the fraction of the series observed.
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 presents XMTC, an interactive visual analytics tool for early classification of multivariate time series, applied to reach-to-grasp hand kinematics. The approach trains DrCIF models on prefixes of increasing length (moving window with step 10) and provides coordinated views—a temporal accuracy plot, confusion matrix heatmaps, per-series prediction heatmaps, and partial dependence plots—to let analysts choose an early-vs-accurate trade-off and identify difficult classes and influential features. Three real motion-capture experiments are analyzed; in Experiment 3 the authors report 73.02% accuracy at window size 360 (before any object is grasped) and 91.01% at window size 640, with qualitative explanations of confusable classes and feature impacts. A leave-one-user-out test on Experiment 1 is reported as a generalizability check.

Significance. If the accuracy claims held under deployment conditions, the paper would make a useful contribution: it packages an existing strong MTSC classifier (DrCIF) into a visually guided early-prediction workflow, with post-hoc, model-agnostic explanations that transfer to other classifiers. Strengths include real high-frequency motion data, multiple controlled experiments, and the explicit attempt to measure cross-user generalization in Section 6.4. The main methodological question is whether the headline before-grasp accuracies, obtained from a random trial-level split, reflect performance for unseen users; the paper's own leave-one-out numbers show a substantial drop, so the generalizability claim is currently not established for the central case study.

major comments (4)
  1. [Sections 6.1, 6.3, 6.4] The evaluation protocol is the load-bearing issue. Section 6.1 states that training and testing data are 'randomly selected and stratified'; with 29 participants and two trials each, a trial-level split places the same user's kinematics in both training and test sets, and DrCIF can exploit per-user motion style rather than only object-intention signals. The recommended window-380 model for Experiment 1 achieves 71.09% on this split, but the leave-one-out test in Section 6.4 reports 60.62% (sigma 10.19%, median 62.5%). This 10.5-point drop is material and should not be called 'slight' as in Section 7. The same issue affects Experiment 3, where the headline 73.02% at window 360 and 91.01% at window 640 are reported only on the random split; no leave-one-user-out or other user-independent evaluation is reported for that experiment. The central claim that XMTC can predict the intended object before grasp for unseen users therefore requires a user-independent evaluation for Experiment 3 (and ideally for Experiments 1 and 2), with the recommended trade-off window selected from that evaluation rather than from the random-split test curve.
  2. [Sections 2 and 4] The paper claims that DrCIF received the highest weight among the four HIVE-COTE2 classifiers and that using DrCIF alone yields 'the same accuracy' as HIVE-COTE2, but no accuracy comparison is presented. Since DrCIF is the classifier on which all reported results depend, this equivalence is load-bearing. Please provide HIVE-COTE2 versus DrCIF test accuracies for a representative set of window sizes (including the recommended windows in Section 6), or state precisely the criterion used to conclude 'without loss of accuracy'.
  3. [Section 4] The stretching of short time series to the required window length is specified only as 'we stretch too short time series accordingly for the training phase' (Section 4). The interpolation or padding method is not described, and it is not stated whether the same stretching is applied to test series and in the leave-one-out test. This matters for early-prediction validity, since the procedure must not introduce information from the future or artificially distort the interval features that DrCIF extracts. Please specify the method and justify that it preserves the early-prediction semantics.
  4. [Table 2] Table 2 contains an impossible count: the l cup row reports a total of 24 test samples but 25 correct predictions for Model 1436. This suggests either a data error or inconsistent test sets across window sizes. Since the accuracy tables are the quantitative basis for the findings, the inconsistency must be resolved and the tables checked for similar issues.
minor comments (6)
  1. [Section 7] There is a typo, 'classsification', in the final paragraph of Section 7.
  2. [Section 2] The text uses 'Shapely values' but the correct term is 'Shapley values'; also 'Partial Dependency Plots' should be 'Partial Dependence Plots' for consistency with Section 5.
  3. [Figures 6 and 12] The histograms in the accuracy plots have no y-axis label; adding a label such as 'number of time series' would improve interpretability.
  4. [Figures 16 and 17] The text refers to 'time step' when describing the x-axis of the leave-one-out plots, but the x-axis is the window size; please unify the terminology.
  5. [Section 6.3] The phrase 'two random time series from the class l dice' should read 'two randomly selected time series'.
  6. [Section 6.1] The text first identifies 'r cup and r bottle' as the most difficult classes (based on Table 2) but later says 'r cup and r knife were identified as the most challenging classes'; please reconcile this inconsistency.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: XMTC's claims are empirical measurements of a standard external classifier plus a visualization tool, and its self-citations are non-load-bearing background.

full rationale

The paper contains no equation-level derivation whose output could reduce to its input. The load-bearing quantitative claims are measured test-set accuracies of DrCIF, a standard external ensemble classifier, applied to increasing time-window prefixes of the hand-kinematics data; the reported values (e.g., 71.9% at window 380 for Experiment 1 and 73.02% at window 360 for Experiment 3) are empirical evaluation results, not quantities defined in terms of the conclusion. The tool XMTC adds coordinated visualizations (accuracy plots, confusion matrices, heatmaps, PDPs) whose contents are computed from the trained model and data, so no fitted parameter is renamed as a prediction. The only self-citations are [29] and [37], earlier works by the same group; [29] is explicitly described as having failed to reliably discriminate the synthetic object, so it is not used as authoritative support for the present accuracy claims, and [37] is background on hand-posture prediction. The recommended window is selected from the random-split test-accuracy curve, and the paper's own leave-one-out test shows a lower user-independent accuracy (60.62% vs. 71.09% at window 380), but this is a validation-protocol risk and not circularity: the reported numbers are observed accuracies, and no derivation step equates a prediction to an input by construction. Because the central claims are self-contained empirical evaluations against external data and a standard classifier, the circularity score is 0.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim depends on the chosen window discretization, the unspecified stretching of short series, the trust in DrCIF's default configuration, and the assumption that a random trial-level split approximates user-level generalization. None of these are derived; they are inputs from prior work or ad hoc choices. No new physical or computational entities are introduced.

free parameters (3)
  • window step size = 10 time steps
    Windows grow in steps of 10; the accuracy curve and the reported trade-off point depend on this discretization (Section 4).
  • initial window size = 10
    The earliest model is trained on [0,10]; changing this shifts the whole accuracy curve (Section 4).
  • train/test split ratio = 80/20
    All experiments use one 80/20 stratified split; no multiple seeds or confidence intervals are reported (Section 6.1).
assumptions (4)
  • domain assumption A random 80/20 split across trials measures real-world generalization
    Section 6.1 uses a random stratified split; Section 6.4 shows a 10-point drop in leave-one-out, so the assumption is questionable.
  • ad hoc to paper Short time series can be stretched to the required window length without harming early-prediction validity
    Section 4 says 'we stretch too short time series accordingly for the training phase' but gives no method, so the assumption is unverified.
  • domain assumption DrCIF with default parameters is an appropriate base classifier
    The paper trusts the prior benchmark [11] and does not compare against other classifiers on this data.
  • domain assumption The 12 aperture features suffice to discriminate intended objects
    The feature set is based on HCI domain knowledge (e.g., [31,32]); no ablation against raw sensor data is provided.

how reviews work

0 comments
Cite this review

Pith. "Pith review of XMTC: Explainable Early Classification of Multivariate Time Series in Reach-to-Grasp Hand Kinematics." pith.science (2026). https://pith.science/paper/CJBB4UVF

@misc{pith2026250204398,
  author       = {Pith},
  title        = {Pith review of: XMTC: Explainable Early Classification of Multivariate Time Series in Reach-to-Grasp Hand Kinematics},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CJBB4UVF}},
  note         = {Machine review of arXiv:2502.04398}
}
read the original abstract

Hand kinematics can be measured in Human-Computer Interaction (HCI) with the intention to predict the user's intention in a reach-to-grasp action. Using multiple hand sensors, multivariate time series data are being captured. Given a number of possible actions on a number of objects, the goal is to classify the multivariate time series data, where the class shall be predicted as early as possible. Many machine-learning methods have been developed for such classification tasks, where different approaches produce favorable solutions on different data sets. We, therefore, employ an ensemble approach that includes and weights different approaches. To provide a trustworthy classification production, we present the XMTC tool that incorporates coordinated multiple-view visualizations to analyze the predictions. Temporal accuracy plots, confusion matrix heatmaps, temporal confidence heatmaps, and partial dependence plots allow for the identification of the best trade-off between early prediction and prediction quality, the detection and analysis of challenging classification conditions, and the investigation of the prediction evolution in an overview and detail manner. We employ XMTC to real-world HCI data in multiple scenarios and show that good classification predictions can be achieved early on with our classifier as well as which conditions are easy to distinguish, which multivariate time series measurements impose challenges, and which features have most impact.

Figures

Figures reproduced from arXiv: 2502.04398 by the authors.

Figure 1
Figure 1. Weights (y-axis) of the four classifiers in HIVE [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Overview of the XMTC tool, showing the user interface elements of the settings panel (a, b, c) and visualizations [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Overview of XMTC tool (continuation of Figure 2). Users can select a model in the setting panel(c) and observe the [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (16 more)
Figure 4
Figure 4. Figure 4: Hovering information of the accuracy plot. When [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Additionally, a vertical dashed line is shown to help [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 5
Figure 5. Figure 5: Hovering information of the heatmap: The window size and the probability predictions of the model trained on the [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: Accuracy plot of the test dataset for Experiment 1, starting at 35 [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]
Figure 7
Figure 7. Figure 7: Confusion matrix and accuracy of the four models [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]
Figure 8
Figure 8. Figure 8: Models’ confidence over time for two samples in the [PITH_FULL_IMAGE:figures/full_fig_p011_8.png]
Figure 9
Figure 9. Figure 9: Models’ confidence over time for two samples in the [PITH_FULL_IMAGE:figures/full_fig_p012_9.png]
Figure 10
Figure 10. Figure 10: Partial Dependence Plot for two challenging classes, [PITH_FULL_IMAGE:figures/full_fig_p013_10.png]
Figure 11
Figure 11. Figure 11: Accuracy of test dataset for Experiment 2. Achieving an initial accuracy of 22 [PITH_FULL_IMAGE:figures/full_fig_p014_11.png]
Figure 12
Figure 12. Figure 12: Accuracy plot of the test dataset for Experiment 3. The model initially achieves an accuracy of 29 [PITH_FULL_IMAGE:figures/full_fig_p014_12.png]
Figure 13
Figure 13. Figure 13: Confusion matrix and accuracy of the four models [PITH_FULL_IMAGE:figures/full_fig_p015_13.png]
Figure 14
Figure 14. Figure 14: Model’s confidence over time for two samples in the [PITH_FULL_IMAGE:figures/full_fig_p016_14.png]
Figure 15
Figure 15. Figure 15: Partial Dependence Plots of the l dice class for two time steps (360 and 1378) in Experiment 3. The model’s accuracy at these time steps is 74% and 100%, respectively. The features tiax and tlax exhibit very similar patterns, while tmax and trax show noticeable change…
Figure 16
Figure 16. Figure 16: Bar chart depicting the average accuracy and standard deviation for the leave-one-out test in Experiment 1. After [PITH_FULL_IMAGE:figures/full_fig_p017_16.png]
Figure 17
Figure 17. Figure 17: Box plot illustrating the accuracy distribution of each model in the leave-one-out test for Experiment 1. The proposed [PITH_FULL_IMAGE:figures/full_fig_p018_17.png]
Figure 18
Figure 18. Figure 18: Effect of omitting individual users on model performance in the leave-one-out test for Experiment 1. The plot shows [PITH_FULL_IMAGE:figures/full_fig_p018_18.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 41 canonical work pages

  1. [1]

    Helena: Real-time contact-free monitoring of sleep ac- tivities and events around the bed,

    J. Clemente, M. Valero, F. Li, C. Wang, and W. Song, “Helena: Real-time contact-free monitoring of sleep ac- tivities and events around the bed,” in2020 IEEE Inter- national Conference on Pervasive Computing and Com- munications (PerCom). IEEE, 2020, pp. 1–10

    work page 2020

  2. [2]

    A predic- tion approach for stock market volatility based on time series data,

    S. M. Idrees, M. A. Alam, and P. Agarwal, “A predic- tion approach for stock market volatility based on time series data,” IEEE Access , vol. 7, pp. 17 287–17 298, 2019

    work page 2019

  3. [3]

    A comparison of time series models for forecasting gdp growth and inflation,

    M. Marcellino, “A comparison of time series models for forecasting gdp growth and inflation,” Bocconi Univer- sity, Italia, 2007

    work page 2007

  4. [4]

    Photovoltaic power forecasting based on ga improved bi-lstm in microgrid without meteorological informa- tion,

    H. Zhen, D. Niu, K. Wang, Y . Shi, Z. Ji, and X. Xu, “Photovoltaic power forecasting based on ga improved bi-lstm in microgrid without meteorological informa- tion,” Energy, vol. 231, p. 120908, 2021. 12 Figure 10: Partial Dependence Plot for two challenging classes,r cup and r knife, in Experiment 1. Features such as tiay, tmay, and tray exhibit similar ...

    work page 2021

  5. [5]

    Transductive lstm for time-series prediction: An application to weather fore- casting,

    Z. Karevan and J. A. Suykens, “Transductive lstm for time-series prediction: An application to weather fore- casting,” Neural Networks, vol. 125, pp. 1–9, 2020

    work page 2020

  6. [6]

    Children age group detection based on human–computer interaction and time series analysis,

    J. C. Ruiz-Garcia, C. Hojas, R. Tolosana, R. Vera- Rodriguez, A. Morales, J. Fierrez, J. Ortega-Garcia, and J. Herreros-Rodriguez, “Children age group detection based on human–computer interaction and time series analysis,” International Journal on Document Analysis and Recognition (IJDAR), pp. 1–11, 2024

    work page 2024

  7. [7]

    Extract: Strong examples from weakly-labeled sensor data,

    D. W. Blalock and J. V . Guttag, “Extract: Strong examples from weakly-labeled sensor data,” in 2016 IEEE 16th International Conference on Data Mining (ICDM). IEEE, 2016, pp. 799–804

    work page 2016

  8. [8]

    Feature extraction tech- niques for human-computer interaction,

    A. Subasi and S. M. Qaisar, “Feature extraction tech- niques for human-computer interaction,” in Artifi- cial Intelligence and Multimodal Signal Processing in Human-Machine Interaction. Elsevier, 2025, pp. 43– 61

    work page 2025

Show all 43 references
  1. [9]

    A methodological and structural review of hand gesture recognition across diverse data modali- ties,

    J. Shin, A. S. M. Miah, M. H. Kabir, M. A. Rahim, and A. Al Shiam, “A methodological and structural review of hand gesture recognition across diverse data modali- ties,” IEEE Access, 2024

    2024

  2. [10]

    Deep learning for hand gesture recognition in virtual museum using wearable vision sensors,

    N. Zerrouki, F. Harrou, A. Houacine, R. Bouarroudj, M. Y . Cherifi, A.-D. A. Zouina, and Y . Sun, “Deep learning for hand gesture recognition in virtual museum using wearable vision sensors,” IEEE Sensors Journal, 2024

    2024

  3. [11]

    Middlehurst, J

    M. Middlehurst, J. Large, M. Flynn, J. Lines, A. Bostrom, and A. Bagnall, “Hive-cote 2.0: a new 13 Figure 11: Accuracy of test dataset for Experiment 2. Achieving an initial accuracy of 22 .30% and increasing to 59 .71% by time step 630, stabilizing at approximately 60% for th...

    2021

  4. [12]

    Inceptiontime: Finding alexnet for time series classification,

    H. Ismail Fawaz, B. Lucas, G. Forestier, C. Pelletier, D. F. Schmidt, J. Weber, G. I. Webb, L. Idoumghar, P.-A. Muller, and F. Petitjean, “Inceptiontime: Finding alexnet for time series classification,” Data Mining and Knowledge Discovery, vol. 34, no. 6, pp. 1936–1962, 2020

    1936

  5. [13]

    Multirocket: multiple pooling operators and transfor- mations for fast and effective time series classification,

    C. W. Tan, A. Dempster, C. Bergmeir, and G. I. Webb, “Multirocket: multiple pooling operators and transfor- mations for fast and effective time series classification,” Data Mining and Knowledge Discovery, vol. 36, no. 5, pp. 1623–1646, 2022

    2022

  6. [14]

    Deep learning for time series classification and extrinsic regression: A current survey,

    N. Mohammadi Foumani, L. Miller, C. W. Tan, G. I. Webb, G. Forestier, and M. Salehi, “Deep learning for time series classification and extrinsic regression: A current survey,” ACM Computing Surveys , vol. 56, no. 9, pp. 1–45, 2024

    2024

  7. [15]

    Transformers in time series: a survey,

    Q. Wen, T. Zhou, C. Zhang, W. Chen, Z. Ma, J. Yan, and L. Sun, “Transformers in time series: a survey,” in Pro- ceedings of the Thirty-Second International Joint Con- ference on Artificial Intelligence, 2023, pp. 6778–6786

    2023

  8. [16]

    Transformers in time- series analysis: A tutorial,

    S. Ahmed, I. E. Nielsen, A. Tripathi, S. Siddiqui, R. P. Ramachandran, and G. Rasool, “Transformers in time- series analysis: A tutorial,” Circuits, Systems, and Sig- nal Processing, vol. 42, no. 12, pp. 7433–7466, 2023

    2023

  9. [17]

    At- tend and diagnose: Clinical time series analysis using attention models,

    H. Song, D. Rajan, J. Thiagarajan, and A. Spanias, “At- tend and diagnose: Clinical time series analysis using attention models,” in Proceedings of the AAAI confer- ence on artificial intelligence, vol. 32, no. 1, 2018

    2018

  10. [18]

    Fast, ac- curate and explainable time series classification through randomization,

    N. Cabello, E. Naghizade, J. Qi, and L. Kulik, “Fast, ac- curate and explainable time series classification through randomization,” Data Mining and Knowledge Discov- ery, vol. 38, no. 2, pp. 748–811, 2024

    2024

  11. [19]

    The great multivariate time series classifi- cation bake off: a review and experimental evaluation of recent algorithmic advances,

    A. P. Ruiz, M. Flynn, J. Large, M. Middlehurst, and A. Bagnall, “The great multivariate time series classifi- cation bake off: a review and experimental evaluation of recent algorithmic advances,” Data Mining and Knowl- edge Discovery, vol. 35, no. 2, pp. 401–449, 2021. 14 Lef...

    2021

  12. [20]

    Kamath and J

    U. Kamath and J. Liu, Explainable artificial intelli- gence: an introduction to interpretable machine learn- ing. Springer, 2021, vol. 2

    2021

  13. [21]

    Explainable ai (xai): Core ideas, techniques, and so- lutions,

    R. Dwivedi, D. Dave, H. Naik, S. Singhal, R. Omer, P. Patel, B. Qian, Z. Wen, T. Shah, G. Morgan et al., “Explainable ai (xai): Core ideas, techniques, and so- lutions,” ACM Computing Surveys , vol. 55, no. 9, pp. 1–33, 2023

    2023

  14. [22]

    Survey on ex- plainable ai: techniques, challenges and open issues,

    A. Abusitta, M. Q. Li, and B. C. Fung, “Survey on ex- plainable ai: techniques, challenges and open issues,” Expert Systems with Applications, vol. 255, p. 124710, 2024

    2024

  15. [23]

    Interpretation of time-series deep models: A survey,

    Z. Zhao, Y . Shi, S. Wu, F. Yang, W. Song, and N. Liu, “Interpretation of time-series deep models: A survey,” arXiv preprint arXiv:2305.14582, 2023

    2023 arXiv

  16. [24]

    Peeking inside the black- box: a survey on explainable artificial intelligence (xai),

    A. Adadi and M. Berrada, “Peeking inside the black- box: a survey on explainable artificial intelligence (xai),” IEEE access, vol. 6, pp. 52 138–52 160, 2018

    2018

  17. [25]

    Explainable ai for time series classification: a review, taxonomy and research directions,

    A. Theissler, F. Spinnato, U. Schlegel, and R. Guidotti, “Explainable ai for time series classification: a review, taxonomy and research directions,”Ieee Access, vol. 10, pp. 100 700–100 724, 2022

    2022

  18. [26]

    Efficient and precise interactive hand tracking through joint, continuous optimization of pose and cor- respondences,

    J. Taylor, L. Bordeaux, T. Cashman, B. Corish, C. Ke- skin, T. Sharp, E. Soto, D. Sweeney, J. Valentin, B. Luff et al., “Efficient and precise interactive hand tracking through joint, continuous optimization of pose and cor- respondences,” ACM Transactions on Graphics (ToG), v...

    2016

  19. [27]

    Survey on 3d hand gesture recognition,

    H. Cheng, L. Yang, and Z. Liu, “Survey on 3d hand gesture recognition,” IEEE transactions on circuits and systems for video technology, vol. 26, no. 9, pp. 1659– 1673, 2015

    2015

  20. [28]

    Finexus: Track- ing precise motions of multiple fingertips using mag- netic sensing,

    K.-Y . Chen, S. N. Patel, and S. Keller, “Finexus: Track- ing precise motions of multiple fingertips using mag- netic sensing,” in Proceedings of the 2016 CHI Confer- ence on Human Factors in Computing Systems , 2016, pp. 1504–1514

    2016

  21. [29]

    Reach prediction using finger motion dynamics,

    D. Valkov, P. Kockwelp, F. Daiber, and A. Kr ¨uger, “Reach prediction using finger motion dynamics,” in Extended Abstracts of the 2023 CHI Conference on Hu- man Factors in Computing Systems, 2023, pp. 1–8

    2023

  22. [30]

    A systematic review of human–computer interaction and explainable artificial intelligence in healthcare with arti- ficial intelligence techniques,

    M. Nazar, M. M. Alam, E. Yafi, and M. M. Su’ud, “A systematic review of human–computer interaction and explainable artificial intelligence in healthcare with arti- ficial intelligence techniques,” IEEE Access, vol. 9, pp. 153 316–153 348, 2021

    2021

  23. [31]

    Reach-to-grasp movements: a multimodal techniques study,

    S. Betti, G. Zani, S. Guerra, U. Castiello, and L. Sartori, “Reach-to-grasp movements: a multimodal techniques study,”Frontiers in psychology, vol. 9, p. 990, 2018

    2018

  24. [32]

    Shaping of reach-to-grasp kinematics by intentions: A meta-analysis,

    I. Egmose and S. Køppe, “Shaping of reach-to-grasp kinematics by intentions: A meta-analysis,” Journal of Motor Behavior, vol. 50, no. 2, pp. 155–165, 2018

    2018

  25. [33]

    Coordination between the transport and the grasp components during prehen- sion movements,

    S. Chieffi and M. Gentilucci, “Coordination between the transport and the grasp components during prehen- sion movements,”Experimental brain research, vol. 94, pp. 471–477, 1993

    1993

  26. [34]

    Predicting object size from hand kine- matics: a temporal perspective,

    C. Ansuini, A. Cavallo, A. Koul, M. Jacono, Y . Yang, and C. Becchio, “Predicting object size from hand kine- matics: a temporal perspective,” PLoS One , vol. 10, no. 3, p. e0120432, 2015. 15 TS6 TS38 (a) (b) Figure 14: Model’s confidence over time for two samples in the l dic...

    2015

  27. [35]

    Gradual molding of the hand to object contours,

    M. Santello and J. F. Soechting, “Gradual molding of the hand to object contours,” Journal of neurophysiol- ogy, vol. 79, no. 3, pp. 1307–1320, 1998

    1998

  28. [36]

    A neural model of spatio temporal coordination in pre- hension,

    J. Molina-Vilaplana, J. F. Batlle, and J. L. Coronado, “A neural model of spatio temporal coordination in pre- hension,” in Artificial Neural Networks—ICANN 2002: International Conference Madrid, Spain, August 28–30, 2002 Proceedings 12. Springer, 2002, pp. 9–14

    2002

  29. [37]

    Towards object prediction based on hand postures for reach to grasp interaction,

    F. Daiber, D. Valkov, F. Steinicke, K. H. Hinrichs, and A. Kr ¨uger, “Towards object prediction based on hand postures for reach to grasp interaction,” in Proceed- ings of the ACM CHI Workshop on The 3rd Dimension of CHI: Touching and Designing 3D User Interfaces (3DCHI), 2012

    2012

  30. [38]

    Bake off redux: a review and experimental evaluation of recent time series classification algorithms,

    M. Middlehurst, P. Sch ¨afer, and A. Bagnall, “Bake off redux: a review and experimental evaluation of recent time series classification algorithms,” Data Mining and Knowledge Discovery, pp. 1–74, 2024

    2024

  31. [39]

    Greedy function approximation: a gra- dient boosting machine,

    J. H. Friedman, “Greedy function approximation: a gra- dient boosting machine,”Annals of statistics, pp. 1189– 1232, 2001

    2001

  32. [40]

    A visual analytics conceptual framework for explorable and steerable partial dependence analysis,

    M. Angelini, G. Blasilli, S. Lenti, and G. Santucci, “A visual analytics conceptual framework for explorable and steerable partial dependence analysis,”IEEE Trans- actions on Visualization and Computer Graphics, 2023

    2023

  33. [41]

    Detoxer: a visual debugging tool with multiscope explanations for temporal multilabel classi- fication,

    M. Nourani, C. Roy, D. R. Honeycutt, E. D. Ragan, and V . Gogate, “Detoxer: a visual debugging tool with multiscope explanations for temporal multilabel classi- fication,” IEEE Computer Graphics and Applications , vol. 42, no. 6, pp. 37–46, 2022. 16 Figure 16: Bar chart depict...

    2022

  34. [42]

    catch22: Canonical time- series characteristics: Selected through highly compar- ative time-series analysis,

    C. H. Lubba, S. S. Sethi, P. Knaute, S. R. Schultz, B. D. Fulcher, and N. S. Jones, “catch22: Canonical time- series characteristics: Selected through highly compar- ative time-series analysis,” Data Mining and Knowl- edge Discovery, vol. 33, no. 6, pp. 1821–1852, 2019. 17 Fig...

    2019

  35. [1378]

    (%) l bar 45 13 39 41 41 29% 87% 91% 91% l box 46 14 24 43 40 30% 52% 93% 87% l dice 46 11 34 46 46 24% 74% 100% 100% l plank 46 6 22 35 41 13% 48% 76% 89% r bar 46 26 38 42 43 57% 83% 91% 93% r box 46 7 31 43 45 15% 67% 93% 98% r dice 46 16 40 43 45 35% 87% 93% 98% r plank 46...

Pith tools

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