Pith. sign in

REVIEW 4 major objections 10 minor 27 references

Emotion Recognition in Contemporary Dance Performances Using Laban Movement Analysis

T0 review · 4 major / 10 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read A 54-feature Laban descriptor set reads 12 dance emotions at 96.85% accuracy.

desk verdict Useful LMA feature recipe for dance emotion recognition, but the headline 96.85% accuracy is unsupported as reported because overlapping sliding windows almost certainly leak across folds. read the letter →

arxiv 2504.21154 v1 pith:NL55YEDG submitted 2025-04-29 cs.CV cs.AI

Muhammad Turab , Philippe Colantoni , Damien Muselet , Alain Tremeau This is my paper · ORCID
classification cs.CVcs.AI
keywords emotionrecognitioncontemporarydanceLabanMovementAnalysis3DbodyposeslidingwindowfeaturesrandomforestSVMSHAPexplainability
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 aims to show that emotion recognition in contemporary dance can be handled by handcrafted, explainable movement descriptors rather than opaque deep models. It builds a 54-feature descriptor set from the four components of Laban Movement Analysis—body, effort, shape, and space—and adds temporal dynamics through sliding windows over 3D joint keypoints. On the Dance Motion Capture dataset, with five dancers expressing 12 emotions, the Random Forest classifier reaches 96.85% accuracy and the SVM reaches 93.90%, outperforming the compared baselines. The authors also use SHAP, a game-theory-based feature attribution method, to show which descriptors drive predictions, arguing that the same features can support performance analysis, dance training, and human–computer interaction.

What carries the argument

The load-bearing object is the 54-component LMA descriptor vector, computed per 25-frame sliding window from 3D body keypoints. Body descriptors include joint distances and angles, plus an initiation descriptor that flags which joint starts a movement by comparing displacement over a window against a data-driven threshold $\tau$. Effort descriptors cover Space (windowed path relative to total displacement), Weight (kinetic energy $\sum \frac{1}{2} \alpha_j v_j^2$), and Time (mean acceleration over the window). Shape is represented by body volume from a convex hull, and Space is represented by spatial dispersion and trajectory curvature. Joint weights $\alpha_j$ emphasize the extremities. The resulting vector is fed to Random Forest and SVM classifiers, and SHAP attributes each feature's contribution to individual predictions.

What would settle it

Split the Dance Motion Capture data by whole performance or by dancer so no clip from a performance appears in both training and test folds, retrain the same descriptors and classifiers, and compare the resulting accuracy; if it falls well below 96.85% (or toward chance), the high figure came from near-duplicate frames appearing in both training and testing, not from genuine emotion recognition.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central claim is that adding temporal dynamics to Laban Movement Analysis descriptors yields a compact, interpretable representation of expressive movement that classifies emotional state more accurately than earlier LMA-based methods. The Random Forest on the 25-frame sliding-window descriptor vector reaches 96.85% accuracy across 12 emotion classes; the SVM reaches 93.90%. Individual emotion classes stay roughly balanced, unlike earlier methods where some emotions were recognized much better than others. SHAP analysis identifies body volume and effort time as the strongest contributors, with high values pushing predictions toward expansive, high-energy emotions such as anger or excitement, and low values toward contracted, low-energy states such as sadness or tiredness.

Load-bearing premise

The 3-fold cross-validation treats 25-frame windows as independent samples even though consecutive windows overlap, so the 96.85% accuracy may be measuring memorization of near-duplicate frames from the same performance rather than generalization to unseen dancers or performances.

Editorial extensions

If this is right

  • If the 96.85% figure holds, handcrafted LMA descriptors can match or beat deep-learning classifiers on emotion recognition from skeletal pose while needing far less data.
  • Because emotional state can be read from a 25-frame sliding window, the descriptor vector is usable for real-time emotion feedback in dance training and interactive performance systems.
  • The SHAP rankings give an actionable finding: body volume and effort time are the features that separate high-energy states like anger or excitement from low-energy states like sadness or tiredness.
  • The method's balanced per-class results indicate the descriptor set does not collapse distinct emotions into one another, a failure mode noted for earlier LMA features.

Reading between the lines

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

  • A clean split by entire performance or by dancer would settle whether 96.85% is generalization or overlap memorization; the paper does not report such a split.
  • If dancer-independent accuracy stays high, the same descriptors may transfer to other expressive-motion tasks such as actor performance, physical therapy, or interactive robots, but that transfer is not established here.
  • The dominance of body volume and effort time suggests a tiny descriptor subset may carry most of the signal; trimming the 54 features and re-measuring would test whether the extra machinery is needed.
  • The initiation threshold is computed from the full sequence's standard deviation, so an online version would need an adaptive estimate; comparing offline and online thresholds is a natural next experiment.
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 / 10 minor

Summary. This paper proposes an LMA-based feature extraction framework for emotion recognition from 3D body keypoints in contemporary dance videos. The authors improve existing LMA descriptors with temporal dynamics through a sliding-window approach, introduce new descriptors for the Body, Effort, Space, and Shape components, and train SVM and Random Forest classifiers. They report a highest accuracy of 96.85% (Random Forest) on the Dance Motion Capture dataset with 12 emotions, outperforming several prior methods, and they also use SHAP to analyze feature importance. The main claim is that the proposed descriptors plus a sliding-window scheme yield state-of-the-art emotion recognition accuracy on this dataset.

Significance. If the reported accuracy were obtained under a sound protocol, the paper would be a useful contribution to affective computing and dance analysis: it introduces several novel LMA-inspired descriptors and provides an interpretability analysis via SHAP, which is relatively rare in this area. However, the current evaluation protocol appears to suffer from data leakage due to overlapping windows in cross-validation, which undermines the headline quantitative claim. The comparison with prior work is also not controlled. The strengths are the descriptor design and the explainability analysis, but the central numerical result needs to be re-established.

major comments (4)
  1. [§3.4] The cross-validation protocol is not described in enough detail to rule out data leakage. The paper uses 3-fold cross-validation on 25-frame sliding windows, but it does not state that the folds are split by dancer or performance, nor does it impose a temporal gap between train and test windows. Since consecutive windows overlap by up to 24 frames, a random window-level split would place near-duplicate examples in both training and test folds, so the reported 96.85% accuracy could largely reflect memorization of overlapping windows rather than generalization. The authors must specify whether folds are grouped by performance or dancer, or use a temporal split, and report results under that protocol.
  2. [Table 2] The comparison with prior methods is not controlled. The baseline accuracies are taken from published papers ([3], [7], [27]) that may use different data splits, window sizes, and preprocessing. Without re-implementing the baselines under the same protocol, the claim that the proposed method outperforms all existing methods is not established. Additionally, no standard deviations or confidence intervals are reported for any of the accuracies, making it impossible to judge whether the differences are statistically significant.
  3. [§3.4] The description of hyperparameter tuning is ambiguous. The text says '3 fold cross-validation to divide the dataset into 3 sets including training, testing, and validation' and that the best hyperparameters were selected using 'validation accuracy.' If the same fold is used for both model selection and final evaluation, the test accuracy is optimistically biased. A nested cross-validation or a separate hold-out set is needed to avoid leakage from hyperparameter tuning into the reported test accuracy.
  4. [Figure 2] The caption claims predictions for 'unseen movement sequences,' but the paper does not specify how these sequences were defined or whether they were excluded from training. This claim is therefore unverifiable. The authors should describe how these sequences were selected and confirm that they were not used in any training or hyperparameter tuning.
minor comments (10)
  1. [Abstract] The sentence 'train multiple classifiers' should be 'train multiple classifiers' (grammar) and 'with highest accuracy' should be 'with the highest accuracy.'
  2. [§3.3, Eq. (1)] The notation is incomplete. Please define P_j(t), Δt, and w, and clarify whether the inequality is evaluated for each frame and aggregated (e.g., as a count) or over the whole window.
  3. [§3.3, Eq. (2)] The denominator ∥P_j(T) - P_j(t_1)∥ can be zero for a stationary joint, leading to division by zero; a small epsilon or fallback should be specified.
  4. [§3.3] The 'spatial dispersion' descriptor is described only textually; a precise formula would help reproducibility.
  5. [§3.3] The definition of joint weights α_j is deferred to reference [13] without explaining how they were set; this is important for reproducibility of Eqs. (3)-(5).
  6. [§3.4] '3 fold cross-validation' should be '3-fold cross-validation' and the description of the three sets is confusing; clarify whether a nested CV was used.
  7. [Table 2] Indicate in the caption whether baseline numbers are re-computed by the authors or taken as reported in the original papers.
  8. [Fig. 3] The plot shows accuracy versus window size but lacks error bars; error bars over cross-validation folds should be added.
  9. [§4] The statement 'It is evident that the proposed method outperforms all other methods' is too strong without statistical significance testing; report fold-level variance and perhaps a McNemar test.
  10. [§3.5] The SHAP analysis is presented as 'in-depth explanation,' but the paper only discusses a few top features; consider discussing potential collinearity among features that may affect SHAP values.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the LMA descriptors are computed from motion data, emotion labels come from external dataset ground truth, and the accuracy claim is an empirical classifier result, not a quantity defined in terms of itself.

full rationale

The paper's chain is self-contained against external data. Features (Equations 1-5 plus the ConvexHull volume, spatial dispersion, and joint weights) are computed solely from 3D keypoint motion data and from constants such as joint weights cited from OpenMMLab; none of the descriptors is defined in terms of the emotion labels. The labels are taken from the Dance Motion Capture dataset ground truth provided by the University of Cyprus. The classifiers (SVM, Random Forest) are trained with those labels, and the reported 96.85% accuracy is a measured test-set performance, not a parameter that was fitted to the test labels. Hyperparameter selection with GridSearch on a validation split is standard model selection and is not a fitted input renamed as a prediction. The paper does not invoke any uniqueness theorem, and it does not rely on load-bearing self-citations: the cited works are prior datasets, pose estimators, feature-survey references, and classifier baselines, not prior papers by the same authors that would smuggle in an ansatz. The only substantive methodological concern in the paper is the evaluation protocol: Section 3.4 describes 3-fold cross-validation on 25-frame sliding windows without stating whether windows are grouped by dancer or performance, so overlapping windows may leak across train/test folds. That is a correctness or generalization risk, not a circularity step: the metric does not reduce by construction to the fit, and the paper gives no equation equating the prediction with an input. Under the hard rule requiring a quoted reduction, no circular step can be exhibited. Therefore the appropriate finding is no significant circularity, score 0.

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

The central claim depends on a handful of tuned constants, a pose-estimation accuracy assumption, an independence assumption for sliding windows, and the LMA-to-emotion mapping. No new physical entities or theoretical objects are introduced, so the ledger is short. The weakest item is the cross-validation independence assumption, which is likely violated by the experimental design.

free parameters (3)
  • Sliding window size w = 25-30 frames
    The window size is selected based on the accuracy curve in Fig. 3, with 25-30 chosen as optimal. It affects every feature descriptor and is load-bearing for the reported accuracy.
  • Initiation threshold tau = standard deviation of sequence
    In Equation 1, tau is a data-driven threshold computed from the standard deviation of the entire joint position sequence. It is an algorithmic choice not independently justified.
  • Joint weights alpha_j = from OpenMMLab [13]
    Used in Equations 3, 4, and 5 to give more importance to extremities. The weights are adopted from prior work rather than fitted to this task, but they are still hand-selected constants affecting the features.
assumptions (3)
  • domain assumption NLF 3D pose estimates are sufficiently accurate for the computed LMA features.
    Section 3.2 relies on Neural Localizer Fields to produce 3D keypoints. Errors in pose estimation propagate into every derived feature, yet no validation of pose accuracy on this dance dataset is provided.
  • domain assumption Sliding windows from the same video are independent samples for cross-validation.
    Section 3.4 describes a 3-fold split but does not state that windows are grouped by dancer or video. Since windows overlap by up to 24 frames, this assumption is likely false and is the main threat to the reported accuracy.
  • domain assumption LMA components capture emotion-relevant movement qualities in contemporary dance.
    The entire feature design rests on the premise that Laban Movement Analysis dimensions (Body, Effort, Shape, Space) correspond to emotional expression. The paper does not independently validate this mapping beyond classification accuracy.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Emotion Recognition in Contemporary Dance Performances Using Laban Movement Analysis." pith.science (2026). https://pith.science/paper/NL55YEDG

@misc{pith2026250421154,
  author       = {Pith},
  title        = {Pith review of: Emotion Recognition in Contemporary Dance Performances Using Laban Movement Analysis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NL55YEDG}},
  note         = {Machine review of arXiv:2504.21154}
}
read the original abstract

This paper presents a novel framework for emotion recognition in contemporary dance by improving existing Laban Movement Analysis (LMA) feature descriptors and introducing robust, novel descriptors that capture both quantitative and qualitative aspects of the movement. Our approach extracts expressive characteristics from 3D keypoints data of professional dancers performing contemporary dance under various emotional states, and trains multiple classifiers, including Random Forests and Support Vector Machines. Additionally, we provide in-depth explanation of features and their impact on model predictions using explainable machine learning methods. Overall, our study improves emotion recognition in contemporary dance and offers promising applications in performance analysis, dance training, and human--computer interaction, with a highest accuracy of 96.85\%.

Figures

Figures reproduced from arXiv: 2504.21154 by the authors.

Figure 1
Figure 1. Overview of the proposed method for Emotion Recognition in Contem [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Ground truth and model predictions for three unseen dance performances. [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Impact of sliding window size on accuracy. [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Impact of top 10 features on models predictions. [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: Impact of features on the model’s predictions for dance movements ex [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: Impact of features on the model’s predictions for dance movements ex [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

27 extracted references · 20 canonical work pages

  1. [3]

    In: Computer Graphics Forum

    Aristidou, A., Charalambous, P., Chrysanthou, Y.: Emotion analysis and classifica- tion: understanding the performers’ emotions using the lma entities. In: Computer Graphics Forum. vol. 34, pp. 262–276. Wiley Online Library (2015)

    work page 2015

  2. [7]

    Computer Animation and Virtual Worlds32(6), e1996 (2021)

    Bai, J., Dai, R., Dai, J., Pan, J.: Emodescriptor: A hybrid feature for emotional classification in dance movements. Computer Animation and Virtual Worlds32(6), e1996 (2021)

    work page 2021

  3. [27]

    IEEE Access8, 124928–124938 (2020)

    Wang, S., Li, J., Cao, T., Wang, H., Tu, P., Li, Y.: Dance emotion recognition based on laban motion analysis using convolutional neural network and long short-term memory. IEEE Access8, 124928–124938 (2020)

    work page 2020

  4. [1]

    Multimedia Tools and Applications 78, 16575–16600 (2019)

    Ajili, I., Ramezanpanah, Z., Mallem, M., Didier, J.Y.: Expressive motions recog- nition and analysis with learning and statistical methods. Multimedia Tools and Applications 78, 16575–16600 (2019)

    work page 2019

  5. [2]

    Knowledge and Information Systems62(8), 2937–2987 (2020)

    Alswaidan, N., Menai, M.E.B.: A survey of state-of-the-art approaches for emotion recognition in text. Knowledge and Information Systems62(8), 2937–2987 (2020)

    work page 2020

  6. [4]

    In: 2014 International Conference on Computer Graph- ics Theory and Applications (GRAPP)

    Aristidou, A., Chrysanthou, Y.: Feature extraction for human motion indexing of acted dance performances. In: 2014 International Conference on Computer Graph- ics Theory and Applications (GRAPP). pp. 1–11. IEEE (2014)

    work page 2014

  7. [5]

    Aristidou, A., Shamir, A., Chrysanthou, Y.: Digital dance ethnography: Organizing large dance collections. J. Comput. Cult. Herit.12(4) (Nov 2019).https://doi. org/10.1145/3344383, https://doi.org/10.1145/3344383

    doi:10.1145/3344383 2019

  8. [6]

    Journal on Computing and Cultural Heritage (JOCCH)8(4), 1–19 (2015)

    Aristidou, A., Stavrakis, E., Charalambous, P., Chrysanthou, Y., Himona, S.L.: Folk dance evaluation using laban movement analysis. Journal on Computing and Cultural Heritage (JOCCH)8(4), 1–19 (2015)

    work page 2015

Show all 27 references
  1. [8]

    The journal of machine learning research13(1), 281–305 (2012)

    Bergstra, J., Bengio, Y.: Random search for hyper-parameter optimization. The journal of machine learning research13(1), 281–305 (2012)

    2012

  2. [9]

    In: The human- computer interaction handbook, pp

    Brave, S., Nass, C.: Emotion in human-computer interaction. In: The human- computer interaction handbook, pp. 103–118. CRC Press (2007)

    2007

  3. [10]

    Machine learning45, 5–32 (2001) Emotion Recognition in Contemporary Dance Performances using LMA 11

    Breiman, L.: Random forests. Machine learning45, 5–32 (2001) Emotion Recognition in Contemporary Dance Performances using LMA 11

    2001

  4. [11]

    Information Sciences582, 593–617 (2022)

    Canal, F.Z., Müller, T.R., Matias, J.C., Scotton, G.G., de Sa Junior, A.R., Pozze- bon, E., Sobieranski, A.C.: A survey on facial emotion recognition techniques: A state-of-the-art literature review. Information Sciences582, 593–617 (2022)

    2022

  5. [12]

    IEEE transactions on pattern analysis and machine intelligence43(1), 172–186 (2019)

    Cao, Z., Hidalgo, G., Simon, T., Wei, S.E., Sheikh, Y.: Openpose: Realtime multi- person 2d pose estimation using part affinity fields. IEEE transactions on pattern analysis and machine intelligence43(1), 172–186 (2019)

    2019

  6. [13]

    https: //github.com/open-mmlab/mmpose (2020)

    Contributors, M.: Openmmlab pose estimation toolbox and benchmark. https: //github.com/open-mmlab/mmpose (2020)

    2020

  7. [14]

    Machine learning20, 273–297 (1995)

    Cortes, C., Vapnik, V.: Support-vector networks. Machine learning20, 273–297 (1995)

    1995

  8. [15]

    Electronic Notes in Theoretical Computer Science343, 35–55 (2019)

    Egger, M., Ley, M., Hanke, S.: Emotion recognition from physiological signal analy- sis: A review. Electronic Notes in Theoretical Computer Science343, 35–55 (2019)

    2019

  9. [16]

    IEEE access 7, 117327–117345 (2019)

    Khalil, R.A., Jones, E., Babar, M.I., Jan, T., Zafar, M.H., Alhussain, T.: Speech emotion recognition using deep learning techniques: A review. IEEE access 7, 117327–117345 (2019)

    2019

  10. [17]

    MacDonald & Evans (1950)

    von Laban, R.: The mastery of movement on the stage. MacDonald & Evans (1950)

    1950

  11. [18]

    In: Proceedings of the 2nd International Workshop on Movement and Computing

    Larboulette, C., Gibet, S.: A review of computable expressive descriptors of human motion. In: Proceedings of the 2nd International Workshop on Movement and Computing. pp. 21–28 (2015)

    2015

  12. [19]

    arXiv preprint arXiv:1906.08172 (2019)

    Lugaresi, C., Tang, J., Nash, H., McClanahan, C., Uboweja, E., Hays, M., Zhang, F., Chang, C.L., Yong, M.G., Lee, J., et al.: Mediapipe: A framework for building perception pipelines. arXiv preprint arXiv:1906.08172 (2019)

    2019 arXiv

  13. [20]

    Nature Machine Intelligence2(1), 2522–5839 (2020)

    Lundberg, S.M., Erion, G., Chen, H., DeGrave, A., Prutkin, J.M., Nair, B., Katz, R., Himmelfarb, J., Bansal, N., Lee, S.I.: From local explanations to global under- standing with explainable ai for trees. Nature Machine Intelligence2(1), 2522–5839 (2020)

    2020

  14. [21]

    In: Guyon, I., Luxburg, U.V., Bengio, S., Wallach, H., Fergus, R., Vishwanathan, S., Garnett, R

    Lundberg, S.M., Lee, S.I.: A unified approach to interpreting model predictions. In: Guyon, I., Luxburg, U.V., Bengio, S., Wallach, H., Fergus, R., Vishwanathan, S., Garnett, R. (eds.) Advances in Neural Information Processing Systems 30, pp. 4765–4774. Curran Associates, Inc....

    2017

  15. [22]

    Procedia Computer Science 176, 390–399 (2020)

    Ramezanpanah, Z., Mallem, M., Davesne, F.: Human action recognition using la- ban movement analysis and dynamic time warping. Procedia Computer Science 176, 390–399 (2020)

    2020

  16. [23]

    arXiv preprint arXiv:2407.07532 (2024)

    Sárándi, I., Pons-Moll, G.: Neural localizer fields for continuous 3d human pose and shape estimation. arXiv preprint arXiv:2407.07532 (2024)

    2024 arXiv

  17. [24]

    PeerJ Computer Science9, e1441 (2023)

    Sun, Q., Wu, X.: A deep learning-based approach for emotional analysis of sports dance. PeerJ Computer Science9, e1441 (2023)

    2023

  18. [25]

    Tan, G., Wang, J.: Research on the mechanism of emotion expression in dance based on machine learning models. J. COMBIN. MATH. COMBIN. COMPUT 124(699), 718 (2025)

    2025

  19. [26]

    Nature methods17(3), 261–272 (2020)

    Virtanen, P., Gommers, R., Oliphant, T.E., Haberland, M., Reddy, T., Courna- peau, D., Burovski, E., Peterson, P., Weckesser, W., Bright, J., et al.: Scipy 1.0: fundamental algorithms for scientific computing in python. Nature methods17(3), 261–272 (2020)

    2020

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.