pith. machine review for the scientific record. sign in

arxiv: 2604.05636 · v1 · submitted 2026-04-07 · 💻 cs.CV

Recognition: 2 theorem links

· Lean Theorem

Towards Athlete Fatigue Assessment from Association Football Videos

Xavier Bou , Nathan Correger , Alexandre Cloots , C\'edric Gavage , Silvio Giancola , C\'edric Schwartz , Fran\c{c}ois Delvaux , Rudi Cloots
show 2 more authors
Marc Van Droogenbroeck Anthony Cioppa
Authors on Pith no claims yet

Pith reviewed 2026-05-10 20:08 UTC · model grok-4.3

classification 💻 cs.CV
keywords fatigue assessmentassociation footballmonocular videoplayer trajectoriesacceleration-speed profilesgame state reconstructionkinematic analysisbroadcast footage
0
0 comments X

The pith

Monocular broadcast videos can recover kinematic patterns compatible with acceleration-speed profiling for fatigue assessment in football.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper investigates whether monocular broadcast videos of association football matches can supply the spatio-temporal data needed for fatigue monitoring. It applies game state reconstruction to derive player trajectories in pitch coordinates and introduces a processing step to generate consistent speed and acceleration values. From these, acceleration-speed profiles are built and tested on benchmark data covering short clips and full halves. The work matters because fatigue links to injury risk and performance, yet current methods rely on intrusive sensors or subjective reports; video offers a passive alternative if the signals prove reliable enough.

Core claim

The paper claims that monocular game state reconstruction can extract player trajectories from broadcast footage and, after kinematic processing, produce acceleration-speed profiles that align with expected patterns for fatigue analysis, even as the method exposes vulnerabilities to noise, calibration issues, and gaps in the video tracks.

What carries the argument

A kinematics processing algorithm that smooths reconstructed player trajectories to yield temporally consistent speed and acceleration estimates in pitch coordinates.

If this is right

  • Acceleration-speed profiles derived from video can act as objective indicators of fatigue-related performance changes.
  • The approach demonstrates feasibility for both brief 30-second segments and extended 45-minute match halves.
  • Broadcast videos provide a low-cost, non-intrusive source for such kinematic analysis.
  • Future work must address trajectory inaccuracies arising from monocular calibration and temporal discontinuities.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • This method could allow real-time fatigue tracking using only existing television broadcasts without additional hardware.
  • It opens possibilities for combining fatigue data with tactical insights from the same video sources.
  • Similar video-based profiling might apply to other team sports with wide broadcast coverage.
  • Improved reconstruction techniques could reduce the observed sensitivity to errors and enable more precise monitoring.

Load-bearing premise

Reconstructed player trajectories from monocular broadcast video are accurate and consistent enough over time to enable reliable acceleration-speed profiling.

What would settle it

A controlled study comparing video-derived acceleration-speed profiles against simultaneous GPS tracking during matches, where the profiles show no matching sensitivity to fatigue states, would disprove the compatibility claim.

Figures

Figures reproduced from arXiv: 2604.05636 by Alexandre Cloots, Anthony Cioppa, C\'edric Gavage, C\'edric Schwartz, Fran\c{c}ois Delvaux, Marc Van Droogenbroeck, Nathan Correger, Rudi Cloots, Silvio Giancola, Xavier Bou.

Figure 1
Figure 1. Figure 1: Overview of the proposed method for athlete fatigue assessment from monocular broadcast football videos. Left: Game state reconstruction enables the recovery of metric player trajectories by projecting detections into pitch coordinates, re￾sulting in time-evolving (x, y) positional signals for each athlete. Right (top): From these trajectories, temporally consistent speed and acceleration signals are estim… view at source ↗
Figure 2
Figure 2. Figure 2: Qualitative examples. Qualitative examples of predicted athlete trajectories (red) and their respective ground-truth trajectories (blue) on the SoccerNet Game State Reconstruction (GSR) test set. (a) Depicts the trajectory of player 17 in the SNGS-200 sequence, and (b) portrays the one of player 3 in the SNGS-199 sequence. As observed, although the predicted trajectories exhibit minor localization noise, t… view at source ↗
Figure 3
Figure 3. Figure 3: Qualitative examples of reliable and non-reliable speed and acceleration predictions. The first two columns depict two athletes with high reliability scores showing strong agreement between predictions and ground truth for both speed and acceleration. The last column shows a low-reliability case where the athlete was not detected for most of the sequence, leading to large error. 0 1 2 3 4 5 6 Speed (m/s) 0… view at source ↗
Figure 4
Figure 4. Figure 4: Comparison between predicted and ground truth A–S profiles for sequence SNGS-199, team left, jersey 3. As can be seen, predicted A–S curve follows the ground truth signal, despite clear fluctuations due to noise and accumulated errors. Green points are the ones selected to regress the linear equation. with caution for the following reasons. First, although temporal correlations are strong, amplitude errors… view at source ↗
Figure 5
Figure 5. Figure 5: Comparison of early (left) and late (right) acceleration–speed (A–S) profiles over a 45-minute broadcast sequence. The early profile shows a gradual decrease in acceleration as speed increases, whereas the late profile indicates a substantially faster decline. diminishes, which is consistent with fatigue effects. Qualitatively, both speed and acceleration signals ex￾hibit high-amplitude fluctuations caused… view at source ↗
read the original abstract

Fatigue monitoring is central in association football due to its links with injury risk and tactical performance. However, objective fatigue-related indicators are commonly derived from subjective self-reported metrics, biomarkers derived from laboratory tests, or, more recently, intrusive sensors such as heart monitors or GPS tracking data. This paper studies whether monocular broadcast videos can provide spatio-temporal signals of sufficient quality to support fatigue-oriented analysis. Building on state-of-the-art Game State Reconstruction methods, we extract player trajectories in pitch coordinates and propose a novel kinematics processing algorithm to obtain temporally consistent speed and acceleration estimates from reconstructed tracks. We then construct acceleration--speed (A-S) profiles from these signals and analyze their behavior as fatigue-related performance indicators. We evaluate the full pipeline on the public SoccerNet-GSR benchmark, considering both 30-second clips and a complete 45-minute half to examine short-term reliability and longer-term temporal consistency. Our results indicate that monocular GSR can recover kinematic patterns that are compatible with A-S profiling while also revealing sensitivity to trajectory noise, calibration errors, and temporal discontinuities inherent to broadcast footage. These findings support monocular broadcast video as a low-cost basis for fatigue analysis and delineate the methodological challenges for future research.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 1 minor

Summary. The manuscript proposes a pipeline for athlete fatigue assessment in association football using monocular broadcast videos. It builds on Game State Reconstruction to extract player trajectories in pitch coordinates, introduces a novel kinematics processing algorithm for temporally consistent speed and acceleration estimates, constructs acceleration-speed (A-S) profiles as fatigue indicators, and evaluates the full pipeline on the SoccerNet-GSR benchmark for both 30-second clips and a full 45-minute half. The central claim is that monocular GSR can recover kinematic patterns compatible with A-S profiling, while revealing sensitivity to trajectory noise, calibration errors, and temporal discontinuities.

Significance. If the compatibility claim is substantiated through quantitative validation against ground-truth data, the work could enable low-cost, non-intrusive fatigue monitoring from existing broadcast footage, complementing or reducing reliance on GPS sensors or lab tests. The use of a public benchmark and explicit acknowledgment of methodological challenges (noise sensitivity, discontinuities) are strengths that support reproducibility and future extensions. However, the current absence of direct anchoring to high-precision tracking or fatigue-labeled data limits immediate applicability and impact.

major comments (2)
  1. [Abstract] Abstract: the claim that monocular GSR recovers 'kinematic patterns that are compatible with A-S profiling' is presented without any quantitative metrics, error bars, profile-shape agreement scores, or comparisons to ground-truth fatigue indicators; evaluation uses only position annotations from SoccerNet-GSR, leaving the compatibility assertion unanchored and vulnerable to reconstruction artifacts.
  2. [Evaluation] Evaluation section: the central claim requires that derived speed/acceleration signals support reliable A-S profiling for fatigue analysis, yet no synchronized GPS, optical tracking, or fatigue-labeled data is used to quantify agreement in high-intensity running volume or fatigue-induced shifts; without such external reference the observed patterns cannot be distinguished from noise or calibration effects.
minor comments (1)
  1. [Abstract] Abstract: the description of the 'novel kinematics processing algorithm' could briefly indicate the key steps (e.g., filtering or interpolation method) to help readers assess how temporal consistency is achieved.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the constructive comments. We address the major comments point by point below, indicating where revisions will be made to the manuscript.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the claim that monocular GSR recovers 'kinematic patterns that are compatible with A-S profiling' is presented without any quantitative metrics, error bars, profile-shape agreement scores, or comparisons to ground-truth fatigue indicators; evaluation uses only position annotations from SoccerNet-GSR, leaving the compatibility assertion unanchored and vulnerable to reconstruction artifacts.

    Authors: We agree that the abstract claim would benefit from stronger quantitative anchoring. The current evaluation derives A-S profiles from position annotations and demonstrates internal consistency (e.g., stable profile shapes in short clips and expected temporal shifts over the 45-minute half), but lacks direct metrics such as profile-shape agreement or error bars relative to external references. We will revise the abstract to qualify the claim as indicating 'qualitatively compatible' patterns while highlighting the documented sensitivities, and we will add binned statistics (mean acceleration per speed interval with standard deviations) to the results to provide quantitative descriptors of the profiles. revision: yes

  2. Referee: [Evaluation] Evaluation section: the central claim requires that derived speed/acceleration signals support reliable A-S profiling for fatigue analysis, yet no synchronized GPS, optical tracking, or fatigue-labeled data is used to quantify agreement in high-intensity running volume or fatigue-induced shifts; without such external reference the observed patterns cannot be distinguished from noise or calibration effects.

    Authors: This observation is correct: the SoccerNet-GSR benchmark supplies only video-derived position annotations, with no synchronized high-precision GPS, optical tracking, or fatigue labels available. Consequently, we cannot compute quantitative agreement metrics for high-intensity running volume or fatigue-induced shifts. Our analysis instead emphasizes internal consistency checks and explicit sensitivity analysis to trajectory noise and calibration errors, as already noted in the manuscript. We will expand the evaluation and discussion sections to include error-propagation analysis from positions to accelerations and a clearer statement of this limitation, but we cannot introduce external ground-truth datasets within the scope of the current work. revision: partial

standing simulated objections not resolved
  • Direct quantitative validation against synchronized GPS, optical tracking, or fatigue-labeled data, since no such references exist in the SoccerNet-GSR benchmark or the present study design.

Circularity Check

0 steps flagged

No significant circularity; empirical pipeline evaluated on public benchmark

full rationale

The paper describes extraction of trajectories via existing GSR methods, a novel kinematics processing algorithm, construction of A-S profiles, and empirical evaluation on SoccerNet-GSR for both short clips and full halves. No equations or derivations are shown that reduce claimed compatibility or sensitivity observations to fitted parameters, self-definitions, or self-citation chains. The central results are presented as observed patterns from the benchmark rather than tautological outputs, with explicit acknowledgment of noise and discontinuities as limitations. This keeps the derivation chain self-contained against external data.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review supplies no explicit free parameters, axioms, or invented entities; the pipeline rests on unstated assumptions about trajectory accuracy that are not quantified here.

pith-pipeline@v0.9.0 · 5549 in / 1044 out tokens · 19031 ms · 2026-05-10T20:08:18.847676+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

  • IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear
    ?
    unclear

    Relation between the paper passage and the cited Recognition theorem.

    We extract player trajectories in pitch coordinates and propose a novel kinematics processing algorithm to obtain temporally consistent speed and acceleration estimates... construct acceleration–speed (A–S) profiles... slope and intercept parameters of the A–S diagram are analyzed as fatigue-related kinematic indicators

  • IndisputableMonolith/Foundation/ArrowOfTime.lean forward_accumulates unclear
    ?
    unclear

    Relation between the paper passage and the cited Recognition theorem.

    temporal smoothing... Kalman filter... Savitzky–Golay filter... l_n temporal window... ln = 20 (1.64 s)

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

88 extracted references · 8 canonical work pages · 1 internal anchor

  1. [1]

    Effect of playing position and microcycle days on the acceleration speed profile of elite football players.Sci

    Antonio Alonso-Callejo, Jorge Garc ´ıa-Unanue, Andr ´es Perez-Guerra, David Gomez, Javier S ´anchez-S´anchez, Leonor Gallardo, Jose Mar ´ıa Oliva-Lozano, and Jose Luis Felipe. Effect of playing position and microcycle days on the acceleration speed profile of elite football players.Sci. Reports, 12(1), 2022. 3

    2022

  2. [2]

    Validity and reliability of the acceleration- speed profile for assessing running kinematics’ variables de- rived from the force-velocity profile in professional soccer players.J

    Antonio Alonso-Callejo, Jorge Garc ´ıa-Unanue, Marc Guitart-Trench, Carlos Majano, Leonor Gallardo, and Jose Luis Felipe. Validity and reliability of the acceleration- speed profile for assessing running kinematics’ variables de- rived from the force-velocity profile in professional soccer players.J. Strength & Cond. Res., 38(3):563–570, 2023. 2, 3

    2023

  3. [3]

    AiCommentator: A multimodal conversational agent for embedded visualization in football viewing

    Peter Andrews, Oda Elise Nordberg, Stephanie Zu- bicueta Portales, Nj ˚al Borch, Frode Guribye, Kazuyuki Fu- jita, and Morten Fjeld. AiCommentator: A multimodal conversational agent for embedded visualization in football viewing. InInt. Conf. Intell. User Interfaces, pages 14–34, Greenville, SC, USA, 2024. ACM. 3

    2024

  4. [4]

    Using player’s body- orientation to model pass feasibility in soccer

    Adri `a Arbu ´es Sang ¨uesa, Adri `an Mart´ın, Javier Fern ´andez, Coloma Ballester, and Gloria Haro. Using player’s body- orientation to model pass feasibility in soccer. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 3875–3884, Seattle, W A, USA, 2020. IEEE. 3

    2020

  5. [5]

    Clausi, and John Zelek

    Bavesh Balaji, Jerrin Bright, Harish Prakash, Yuhao Chen, David A. Clausi, and John Zelek. Jersey number recognition using keyframe identification from low-resolution broadcast videos. InInt. ACM Work. Multimedia Content Anal. Sports (MMSports), pages 123–130, Ottawa, Ontario, Can., 2023. ACM. 2, 3

    2023

  6. [6]

    Assessment of fatigue and recovery in sport: NarrativeReview.Int

    Thomas Bestwick-Stevenson, Rebecca Toone, Emma Neu- pert, Kimberley Edwards, and Stefan Kluzek. Assessment of fatigue and recovery in sport: NarrativeReview.Int. J. Sports Medicine, 43(14):1151–1162, 2022. 1

    2022

  7. [7]

    Exploring the role of sprint biomechanics in hamstring strain injuries: A current opinion on existing concepts and evidence.Sports Medicine, 54(4):783–793, 2023

    Christopher Bramah, Jurdan Mendiguchia, Thomas Dos’Santos, and Jean-Beno¨ıt Morin. Exploring the role of sprint biomechanics in hamstring strain injuries: A current opinion on existing concepts and evidence.Sports Medicine, 54(4):783–793, 2023. 3

    2023

  8. [8]

    Padr ´on, Bernard Ghanem, and Marc Van Droogenbroeck

    Bruno Cabado, Anthony Cioppa, Silvio Giancola, Andr ´es Villa, Bertha Guijarro-Berdi ˜nas, Emilio J. Padr ´on, Bernard Ghanem, and Marc Van Droogenbroeck. Beyond the Pre- mier: Assessing action spotting transfer capability across di- verse domains. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 3386–3398, Seattle, W A, USA, 2024. IEEE. 3

    2024

  9. [9]

    Pedro Cardoso, Francisco Tavares, Nuno Loureiro, Ruben Ferreira, Jo ˜ao Pedro Ara ´ujo, Joana Reis, and Jo ˜ao R. Vaz. In-situ acceleration-speed profile of an elite soccer academy: A cross-sectional study.J. Sports Sci., 41(20):1868–1874,

  10. [10]

    Time- to-injury forecasting in elite female football: A DeepHit sur- vival approach.arXiv, abs/2601.19479, 2026

    Victoria Catterall, Cise Midoglu, and Stephen Lynch. Time- to-injury forecasting in elite female football: A DeepHit sur- vival approach.arXiv, abs/2601.19479, 2026. 1

    work page arXiv 2026

  11. [11]

    A bottom-up approach based on semantics for the interpretation of the main camera stream in soccer games

    Anthony Cioppa, Adrien Deli `ege, and Marc Van Droogen- broeck. A bottom-up approach based on semantics for the interpretation of the main camera stream in soccer games. InIEEE Int. Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), CVsports, pages 1846–1855, Salt Lake City, UT, USA, 2018. 3

    2018

  12. [12]

    ARTHuS: Adaptive real-time human segmentation in sports through online distillation

    Anthony Cioppa, Adrien Deli `ege, Maxime Istasse, Christophe De Vleeschouwer, and Marc Van Droogen- broeck. ARTHuS: Adaptive real-time human segmentation in sports through online distillation. InIEEE Int. Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), CVsports, pages 2505–2514, Long Beach, CA, USA, 2019. IEEE. 3

    2019

  13. [13]

    Moeslund

    Anthony Cioppa, Adrien Deli `ege, Silvio Giancola, Bernard Ghanem, Marc Van Droogenbroeck, Rikke Gade, and Thomas B. Moeslund. A context-aware loss function for ac- tion spotting in soccer videos. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), pages 13123–13133, Seattle, W A, USA, 2020. IEEE. 3

    2020

  14. [14]

    Camera calibration and player local- ization in SoccerNet-v2 and investigation of their represen- tations for action spotting

    Anthony Cioppa, Adrien Deli `ege, Silvio Giancola, Flori- ane Magera, Olivier Barnich, Bernard Ghanem, and Marc Van Droogenbroeck. Camera calibration and player local- ization in SoccerNet-v2 and investigation of their represen- tations for action spotting. InIEEE Int. Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), CVsports, pages 4532– 4541, Nashvil...

    2021

  15. [15]

    Scaling up Soccer- Net with multi-view spatial localization and re-identification

    Anthony Cioppa, Adrien Deli `ege, Silvio Giancola, Bernard Ghanem, and Marc Van Droogenbroeck. Scaling up Soccer- Net with multi-view spatial localization and re-identification. Sci. Data, 9(1):1–9, 2022. 2, 3

    2022

  16. [16]

    SoccerNet-tracking: Multiple object tracking dataset and benchmark in soccer videos

    Anthony Cioppa, Silvio Giancola, Adrien Deli `ege, Le Kang, Xin Zhou, Zhiyu Cheng, Bernard Ghanem, and Marc Van Droogenbroeck. SoccerNet-tracking: Multiple object tracking dataset and benchmark in soccer videos. InIEEE Int. Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), CVsports, pages 3490–3501, New Orleans, LA, USA, 2022. IEEE. 2, 3

    2022

  17. [17]

    Anthony Cioppa, Silvio Giancola, Vladimir Somers, Vic- tor Joos, Floriane Magera, Jan Held, Seyed Abolfazl Ghasemzadeh, Xin Zhou, Karolina Seweryn, Mateusz Kowalczyk, Zuzanna Mr´oz, Szymon Łukasik, Michał Hało´n, Hassan Mkhallati, Adrien Deli `ege, Carlos Hinojosa, Karen Sanchez, Amir M. Mansourian, Pierre Miralles, Olivier Barnich, Christophe De Vleescho...

    work page arXiv 2024

  18. [18]

    Anthony Cioppa, Silvio Giancola, Vladimir Somers, Flori- ane Magera, Xin Zhou, Hassan Mkhallati, Adrien Deli `ege, Jan Held, Carlos Hinojosa, Amir M. Mansourian, Pierre Miralles, Olivier Barnich, Christophe De Vleeschouwer, Alexandre Alahi, Bernard Ghanem, Marc Van Droogen- broeck, Abdullah Kamal, Adrien Maglo, Albert Clap ´es, Amr Abdelaziz, Artur Xarles...

    2023

  19. [19]

    da Silva, ´Alvaro Novillo, Alberto Aleta, Roberto L´opez del Campo, Ricardo Resta, Yamir Moreno, and Javier M

    Marco F. da Silva, ´Alvaro Novillo, Alberto Aleta, Roberto L´opez del Campo, Ricardo Resta, Yamir Moreno, and Javier M. Buld ´u. The interplay between pitch control and top speed in soccer considering the stamina factor.Sci. Re- ports, 15(1):1–12, 2025. 1

    2025

  20. [20]

    Moeslund

    Mohamad Dalal, Artur Xarles, Anthony Cioppa, Silvio Gi- ancola, Marc Van Droogenbroeck, Bernard Ghanem, Albert Clap´es, Sergio Escalera, and Thomas B. Moeslund. Ac- tion anticipation from soccernet football video broadcasts. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 6070–6081, Nashville, TN, USA, 2025. IEEE. 3

    2025

  21. [21]

    Seikavandi, Jacob V

    Adrien Deli `ege, Anthony Cioppa, Silvio Giancola, Meisam J. Seikavandi, Jacob V . Dueholm, Kamal Nas- rollahi, Bernard Ghanem, Thomas B. Moeslund, and Marc Van Droogenbroeck. SoccerNet-v2: A dataset and bench- marks for holistic understanding of broadcast soccer videos. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 4503–4514, Nashv...

    2021

  22. [22]

    Monitoring training load and fatigue in soccer players with physiological markers.Physiol

    L ´eo Djaoui, Monoem Haddad, Karim Chamari, and Alexan- dre Dellal. Monitoring training load and fatigue in soccer players with physiological markers.Physiol. & Behav., 181: 86–94, 2017. 1

    2017

  23. [23]

    Pascal Edouard, Johan Lahti, Luca Fleres, Juha Ahtiainen, Juha-Jaakko Ulvila, Tiitus Lehtinen, Niklas Virtanen, Toni Taipale, Michel Bellver, Ville Peltonen, Max Thibault, Toni Huuhka, Risto-Matti Toivonen, Jean-Benoit Morin, and Ju- rdan Mendiguchia. A musculoskeletal multifactorial indi- vidualised programme for hamstring muscle injury risk re- duction ...

  24. [24]

    Foul prediction with estimated poses from soccer broadcast video.Sports Eng., 28(2):1–10, 2025

    Jiale Fang, Calvin Yeung, and Keisuke Fujii. Foul prediction with estimated poses from soccer broadcast video.Sports Eng., 28(2):1–10, 2025. 3

    2025

  25. [25]

    Frequency distribution of the values of the correlation coefficient in samples from an indefinitely large population.Biometrika, 10(4):507–521, 1915

    Ronald Aylmer Fisher. Frequency distribution of the values of the correlation coefficient in samples from an indefinitely large population.Biometrika, 10(4):507–521, 1915. 5

    1915

  26. [26]

    Soccer game sum- marization using audio commentary, metadata, and captions

    Sushant Gautam, Cise Midoglu, Saeed Shafiee Sabet, Di- nesh Baniya Kshatri, and P ˚al Halvorsen. Soccer game sum- marization using audio commentary, metadata, and captions. InWork. User-centric Narrat. Summ. Long Videos (narsum), pages 13–22, Lisboa Portugal, 2022. ACM. 3

    2022

  27. [27]

    Riegler, P ˚al Halvorsen, and Mubarak Shah

    Sushant Gautam, Mehdi Houshmand Sarkhoosh, Jan Held, Cise Midoglu, Anthony Cioppa, Silvio Giancola, Va- jira Thambawita, Michael A. Riegler, P ˚al Halvorsen, and Mubarak Shah. SoccerNet-Echoes: A soccer game audio commentary dataset.Int. Symp. Multimedia (ISM), pages 71–78, 2024. 3

    2024

  28. [28]

    SoccerNet: A scalable dataset for action spotting in soccer videos

    Silvio Giancola, Mohieddine Amine, Tarek Dghaily, and Bernard Ghanem. SoccerNet: A scalable dataset for action spotting in soccer videos. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 1792–179210, Salt Lake City, UT, USA, 2018. IEEE. 3

    2018

  29. [29]

    Anthony Chan, He Zhu, Hongwei Kan, Jiaming Chu, Jianming Hu, Jianyang Gu, Jin Chen, Jo˜ao V

    Silvio Giancola, Anthony Cioppa, Adrien Deli `ege, Flori- ane Magera, Vladimir Somers, Le Kang, Xin Zhou, Olivier Barnich, Christophe De Vleeschouwer, Alexandre Alahi, Bernard Ghanem, Marc Van Droogenbroeck, Abdulrahman Darwish, Adrien Maglo, Albert Clap ´es, Andreas Luyts, An- drei Boiarov, Artur Xarles, Astrid Orcesi, Avijit Shah, Baoyu Fan, Bharath Com...

    2022

  30. [30]

    Towards active learning for action spotting in association football videos

    Silvio Giancola, Anthony Cioppa, Julia Georgieva, Johsan Billingham, Andreas Serner, Kerry Peek, Bernard Ghanem, and Marc Van Droogenbroeck. Towards active learning for action spotting in association football videos. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 5098–5108, Vancouver, Can., 2023. IEEE. 3

    2023

  31. [31]

    World Sci., 2025

    Silvio Giancola, Anthony Cioppa, Bernard Ghanem, and Marc Van Droogenbroeck.Deep learning for action spotting in association football videos, chapter 2.5, pages 427–459. World Sci., 2025. 2, 3

    2025

  32. [32]

    Silvio Giancola, Anthony Cioppa, Marc Guti ´errez-P´erez, Jan Held, Carlos Hinojosa, Victor Joos, Arnaud Leduc, Floriane Magera, Karen Sanchez, Vladimir Somers, Ar- tur Xarles, Antonio Agudo, Alexandre Alahi, Olivier Bar- nich, Albert Clap ´es, Christophe De Vleeschouwer, Sergio Escalera, Bernard Ghanem, Thomas B. Moeslund, Marc Van Droogenbroeck, Tomoki ...

    work page arXiv 2025

  33. [33]

    Effects of electrical stimu- lation on delayed onset muscle soreness (DOMS): Evidences from laboratory and in-field studies.J

    Maristella Gussoni, Sarah Moretti, Alessandra Vezzoli, Va- lerio Genitoni, Guido Giardini, Costantino Balestra, Gerardo Bosco, Lorenza Pratali, Elisabetta Spagnolo, Michela Mon- torsi, and Simona Mrakic-Sposta. Effects of electrical stimu- lation on delayed onset muscle soreness (DOMS): Evidences from laboratory and in-field studies.J. Funct. Morphol. Ki-...

    2023

  34. [34]

    No bells, just whistles: Sports field registration by leveraging geometric properties

    Marc Guti ´errez-P´erez and Antonio Agudo. No bells, just whistles: Sports field registration by leveraging geometric properties. InIEEE/CVF Conf. Comput. Vis. Pattern Recog- nit. Work. (CVPRW), pages 3325–3334, Seattle, W A, USA,

  35. [35]

    SoccerNet-v3D: Leveraging sports broadcast replays for 3D scene under- standing

    Marc Guti ´errez-P´erez and Antonio Agudo. SoccerNet-v3D: Leveraging sports broadcast replays for 3D scene under- standing. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 5968–5977, Nashville, TN, USA,

  36. [36]

    PnLCalib: Sports field registration via points and lines optimization.Comput

    Marc Guti ´errez-P´erez and Antonio Agudo. PnLCalib: Sports field registration via points and lines optimization.Comput. Vis. Image Underst., 267:104712, 2026. 3

    2026

  37. [37]

    V ARS: Video assistant referee system for automated soc- cer decision making from multiple views

    Jan Held, Anthony Cioppa, Silvio Giancola, Abdullah Hamdi, Bernard Ghanem, and Marc Van Droogenbroeck. V ARS: Video assistant referee system for automated soc- cer decision making from multiple views. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 5086–5097, Vancouver, Can., 2023. IEEE. 3

    2023

  38. [38]

    X-V ARS: Introducing explainability in football refereeing with multi- modal large language models

    Jan Held, Hani Itani, Anthony Cioppa, Silvio Giancola, Bernard Ghanem, and Marc Van Droogenbroeck. X-V ARS: Introducing explainability in football refereeing with multi- modal large language models. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 3267–3279, Seattle, W A, USA, 2024. IEEE

    2024

  39. [39]

    Collins, Umang Bhatt, Bernard Ghanem, and Marc Van Droogenbroeck

    Jan Held, Anthony Cioppa, Silvio Giancola, Elaf Almah- moud, Katherine M. Collins, Umang Bhatt, Bernard Ghanem, and Marc Van Droogenbroeck. Enhancing foot- ball refereeing with AI: V ARS and X-V ARS for assisted decision-making. InMathSport Conference, Luxembourg, 2025

    2025

  40. [40]

    Towards an AI-powered video assistant referee system (V ARS) for association football

    Jan Held, Anthony Cioppa, Silvio Giancola, Abdullah Hamdi, Christel Devue, Bernard Ghanem, and Marc Van Droogenbroeck. Towards an AI-powered video assistant referee system (V ARS) for association football. InScand. Conf. Image Anal. (SCIA), pages 295–309. Springer Nat. Switz., 2025. 3

    2025

  41. [41]

    Pass receiver prediction in soccer using video and players’ trajectories

    Yutaro Honda, Rei Kawakami, Ryota Yoshihashi, Kenta Kato, and Takeshi Naemura. Pass receiver prediction in soccer using video and players’ trajectories. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 3502–3511, New Orleans, LA, USA, 2022. IEEE. 3

    2022

  42. [42]

    Spotting temporally pre- cise, fine-grained events in video

    James Hong, Haotian Zhang, Micha ¨el Gharbi, Matthew Fisher, and Kayvon Fatahalian. Spotting temporally pre- cise, fine-grained events in video. InEur. Conf. Comput. Vis. (ECCV), pages 33–51, Tel Aviv, Isra¨el, 2022. Springer Nat. Switz. 3

    2022

  43. [43]

    Oswald, Marc Polleyfeys, Otmar Hilliges, Manuel Kaufmann, and Jie Song

    Tianjian Jiang, Johsan Billingham, Sebastian M ¨uksch, Juan Zarate, Nicolas Evans, Martin R. Oswald, Marc Polleyfeys, Otmar Hilliges, Manuel Kaufmann, and Jie Song. World- Pose: A world cup dataset for global 3D human pose esti- mation. InEur. Conf. Comput. Vis. (ECCV), pages 343–362. Springer Nat. Switz., 2024. 3

    2024

  44. [44]

    SoccerDB: A large-scale database for com- prehensive video understanding

    Yudong Jiang, Kaixu Cui, Leilei Chen, Canjin Wang, and Changliang Xu. SoccerDB: A large-scale database for com- prehensive video understanding. InInt. ACM Work. Multi- media Content Anal. Sports (MMSports), pages 1–8, Seattle, W A, USA, 2020. ACM. 3

    2020

  45. [45]

    P´arraga-Montilla, Jos´e A

    Pedro Jim ´enez-Reyes, Amador Garcia-Ramos, Juan A. P´arraga-Montilla, Jos´e A. Morcillo-Losa, V ´ıctor Cuadrado- Pe˜nafiel, Adri ´an Casta ˜no-Zambudio, Pierre Samozino, and Jean-Benoˆıt Morin. Seasonal changes in the sprint acceler- ation force-velocity profile of elite male soccer players.J. Strength & Cond. Res., 36(1):70–74, 2020. 4

    2020

  46. [46]

    Rudolf E. Kalman. A new approach to linear filtering and prediction problems.J. Basic Eng., 82(1):35–45, 1960. 4

    1960

  47. [47]

    Mechanisms of exercise-induced muscle damage and fatigue: Intracellular calcium accumu- lation.J

    Yutaka Kano, Takashi Sonobe, Tadakatsu Inagaki, Mizuki Sudo, and David C Poole. Mechanisms of exercise-induced muscle damage and fatigue: Intracellular calcium accumu- lation.J. Phys. Fit. Sports Medicine, 1(3):505–512, 2012. 2

    2012

  48. [48]

    Pixels or Positions? Benchmarking Modalities in Group Activity Recognition

    Drishya Karki, Merey Ramazanova, Anthony Cioppa, Silvio Giancola, and Bernard Ghanem. Pixels or positions? bench- marking modalities in group activity recognition.arXiv, abs/2511.12606, 2025. 3

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  49. [49]

    SoccerNet-Depth: a scalable dataset for monocular depth estimation in sports videos

    Arnaud Leduc, Anthony Cioppa, Silvio Giancola, Bernard Ghanem, and Marc Van Droogenbroeck. SoccerNet-Depth: a scalable dataset for monocular depth estimation in sports videos. InIEEE/CVF Conf. Comput. Vis. Pattern Recog- nit. Work. (CVPRW), pages 3280–3282, Seattle, W A, USA,

  50. [50]

    Jersey number recognition with semi-supervised spa- tial transformer network

    Gen Li, Shikun Xu, Xiang Liu, Lei Li, and Changhu Wang. Jersey number recognition with semi-supervised spa- tial transformer network. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 1864–18647, Salt Lake City, UT, USA, 2018. IEEE. 2

    2018

  51. [51]

    PhD thesis, Universit¨at Augsburg, Germany, 2025

    Katja Ludwig.Human pose estimation in images and videos for sports analytics: 2D keypoint and 3D mesh estimation for challenging scenarios and extreme poses. PhD thesis, Universit¨at Augsburg, Germany, 2025. 3

    2025

  52. [52]

    A universal protocol to bench- mark camera calibration for sports

    Floriane Magera, Thomas Hoyoux, Olivier Barnich, and Marc Van Droogenbroeck. A universal protocol to bench- mark camera calibration for sports. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 3335–3346, Seattle, W A, USA, 2024. IEEE. 2, 3

    2024

  53. [53]

    BroadTrack: Broadcast cam- era tracking for soccer

    Floriane Magera, Thomas Hoyoux, Olivier Barnich, and Marc Van Droogenbroeck. BroadTrack: Broadcast cam- era tracking for soccer. InIEEE/CVF Winter Conf. Appl. Comput. Vis. (WACV), pages 6177–6187, Tucson, AZ, USA,

  54. [54]

    Can geometry save central views for sports field registra- tion? InIEEE/CVF Conf

    Floriane Magera, Thomas Hoyoux, Martin Castin, Olivier Barnich, Anthony Cioppa, and Marc Van Droogenbroeck. Can geometry save central views for sports field registra- tion? InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 6060–6069, Nashville, TN, USA,

  55. [55]

    Effi- cient tracking of team sport players with few game-specific annotations

    Adrien Maglo, Astrid Orcesi, and Quoc-Cuong Pham. Effi- cient tracking of team sport players with few game-specific annotations. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 3460–3470, New Orleans, LA, USA, 2022. IEEE. 2, 3

    2022

  56. [56]

    Individual locating of soccer players from a single moving view.Sensors, 23(18):1–28, 2023

    Adrien Maglo, Astrid Orcesi, Julien Denize, and Quoc Cuong Pham. Individual locating of soccer players from a single moving view.Sensors, 23(18):1–28, 2023. 3

    2023

  57. [57]

    Mansourian, Vladimir Somers, Christophe De Vleeschouwer, and Shohreh Kasaei

    Amir M. Mansourian, Vladimir Somers, Christophe De Vleeschouwer, and Shohreh Kasaei. Multi-task learning for joint re-identification, team affiliation, and role classifica- tion for sports visual tracking. InInt. ACM Work. Multimedia Content Anal. Sports (MMSports), pages 103–112, Ottawa, Ontario, Can., 2023. ACM. 3

    2023

  58. [58]

    Establishing reliable acceleration-speed profiles: Minimum data requirements in rugby union matches.J

    Cl ´ement Maviel, Anthony Couderc, Pascale Duch ´e, Jean- Benoˆıt Morin, and Fabrice Vercruyssen. Establishing reliable acceleration-speed profiles: Minimum data requirements in rugby union matches.J. Sports Sci., 42(24):1–6, 2024. 3

    2024

  59. [59]

    Hicks, Morten Bredsgaard Ran- ders, Ramesh Jain, H ˚avard J

    Cise Midoglu, Andreas Kjæreng Winther, Matthias Boeker, Susann Dahl Pettersen, Sigurd Pedersen, Nourhan Ragab, Tomas Kupka, Steven A. Hicks, Morten Bredsgaard Ran- ders, Ramesh Jain, H ˚avard J. Dagenborg, Svein Arne Pettersen, Dag Johansen, Michael A. Riegler, and P ˚al Halvorsen. A large-scale multivariate soccer athlete health, performance, and positio...

    2024

  60. [60]

    AI- based sports highlight generation for social media

    Cise Midoglu, Saeed Shafiee Sabet, Mehdi Housh- mand Sarkhoosh, Mohammad Majidi, Sushant Gautam, H˚akon Maric Solberg, Tomas Kupka, and P˚al Halvorsen. AI- based sports highlight generation for social media. InPro- ceedings of the Mile-High Video Conference, pages 7–13, Denver, CO, USA, 2024. ACM. 3

    2024

  61. [61]

    SoccerNet-caption: Dense video captioning for soccer broadcasts commentaries

    Hassan Mkhallati, Anthony Cioppa, Silvio Giancola, Bernard Ghanem, and Marc Van Droogenbroeck. SoccerNet-caption: Dense video captioning for soccer broadcasts commentaries. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 5074–5085, Vancouver, Can., 2023. IEEE. 3

    2023

  62. [62]

    di Prampero

    Jean-Benoit Morin, Yann Le Mat, Cristian Osgnach, Andrea Barnab`o, Alessandro Pilati, Pierre Samozino, and Pietro E. di Prampero. Individual acceleration-speed profile in-situ: A proof of concept in professional football players.J. Biomech., 123:110524–110532, 2021. 2, 3, 4, 5

    2021

  63. [63]

    Broad- cast2Pitch: Game state reconstruction from unconstrained soccer videos

    Yin May Oo, Yewon Hwang, Muhammad Amrulloh Rob- bani, Vanyi Chao, Ankhzaya Jamsrandorj, Hoang Quoc Nguyen, Kyung-Ryoul Mun, and Jinwook Kim. Broad- cast2Pitch: Game state reconstruction from unconstrained soccer videos. InIEEE/CVF Winter Conf. Appl. Comput. Vis. (WACV), pages 4483–4493, Tucson, AZ, USA, 2026. 2, 3, 5

    2026

  64. [64]

    Paravlic

    Armin H. Paravlic. Establishing reference values for tensiomyography-derived parameters in soccer players: in- sights from a systematic review, meta-analysis and meta- regression.Biology Sport, 42(1):171–192, 2025. 2

    2025

  65. [65]

    Integrating video tracking and GPS to quantify accelerations and decelerations in elite soccer.Sci

    Eduard Pons, Tom ´as Garc´ıa-Calvo, Francesc Cos, Ricardo Resta, Hugo Blanco, Roberto L´opez del Campo, Jes´us D´ıaz- Garc´ıa, and Juan Jos ´e Pulido-Gonz ´alez. Integrating video tracking and GPS to quantify accelerations and decelerations in elite soccer.Sci. Reports, 11(1):1–10, 2021. 2, 4, 5

    2021

  66. [66]

    Towards universal soccer video under- standing

    Jiayuan Rao, Haoning Wu, Hao Jiang, Ya Zhang, Yanfeng Wang, and Weidi Xie. Towards universal soccer video under- standing. InIEEE/CVF Conf. Comput. Vis. Pattern Recog- nit. (CVPR), pages 8384–8394, Nashville, TN, USA, 2025. IEEE. 3

    2025

  67. [67]

    Rubio-Arias, Vicente ´Avila-Gand´ıa, Fran- cisco Javier L ´opez-Rom´an, Fulgencio Soto-M ´endez, Pe- dro E

    Jacobo ´A. Rubio-Arias, Vicente ´Avila-Gand´ıa, Fran- cisco Javier L ´opez-Rom´an, Fulgencio Soto-M ´endez, Pe- dro E. Alcaraz, and Domingo J. Ramos-Campo. Mus- cle damage and inflammation biomarkers after two ultra- endurance mountain races of different distances: 54 km vs 111 km.Physiol. & Behav., 205:51–57, 2019. 2

    2019

  68. [68]

    Extracting player speed from football videos

    Ole Kristian Rustebakke, Mehdi Houshmand Sarkhoosh, Cise Midoglu, and P ˚al Halvorsen. Extracting player speed from football videos. InInt. Symp. Multimedia (ISM), pages 266–273, Naples, Italy, 2025. IEEE. 3, 8

    2025

  69. [69]

    Santamaria and Kate E

    Luke J. Santamaria and Kate E. Webster. The effect of fa- tigue on lower-limb biomechanics during single-limb land- ings: A systematic review.J. Orthop. & Sports Phys. Ther. (JOSPT), 40(8):464–473, 2010. 1

    2010

  70. [70]

    Riegler, and P ˚al Halvorsen

    Mehdi Houshmand Sarkhoosh, Sushant Gautam, Cise Mi- doglu, Thu Nguyen, Jan Held, Anthony Cioppa, Silvio Gi- ancola, Vajira Thambawita, Michael A. Riegler, and P ˚al Halvorsen. Beyond audio: Enhancing SoccerNet-echoes with multimodal event extraction using LLMs.Int. J. Se- mantic Comput., pages 1–25, 2025. 3

    2025

  71. [71]

    Savitzky and Marcel J

    Abraham. Savitzky and Marcel J. E. Golay. Smoothing and differentiation of data by simplified least squares procedures. Anal. Chem., 36(8):1627–1639, 1964. 4

    1964

  72. [72]

    Saw, Luana C

    Anna E. Saw, Luana C. Main, and Paul B. Gastin. Moni- toring the athlete training response: subjective self-reported measures trump commonly used objective measures: a sys- tematic review.Br. J. Sports Medicine, 50(5):281–291, 2016. 1

    2016

  73. [73]

    SoccerTrack: A dataset and tracking algorithm for soccer with fish-eye and drone videos

    Atom Scott, Ikuma Uchida, Masaki Onishi, Yoshinari Kameda, Kazuhiro Fukui, and Keisuke Fujii. SoccerTrack: A dataset and tracking algorithm for soccer with fish-eye and drone videos. InIEEE/CVF Conf. Comput. Vis. Pattern Recognit. Work. (CVPRW), pages 3568–3578, New Orleans, LA, USA, 2022. IEEE. 3

    2022

  74. [74]

    SoccerTrack v2: A full-pitch multi- view soccer dataset for game state reconstruction.arXiv, abs/2508.01802, 2025

    Atom Scott, Ikuma Uchida, Kento Kuroda, Yufi Kim, and Keisuke Fujii. SoccerTrack v2: A full-pitch multi- view soccer dataset for game state reconstruction.arXiv, abs/2508.01802, 2025. 3

    work page arXiv 2025

  75. [75]

    Improving object detection quality in football through super-resolution techniques

    Karolina Seweryn, Gabriel Che ´c, Szymon Łukasik, and Anna Wr ´oblewska. Improving object detection quality in football through super-resolution techniques. InInt. Conf. Comput. Sci., pages 151–163. Springer Nat. Switz., 2025. 3

    2025

  76. [76]

    Survey of action recognition, spotting, and spatio-temporal localization in soccer — current trends and research perspec- tives.ACM Trans

    Karolina Seweryn, Anna Wr ´oblewska, and Szymon Łukasik. Survey of action recognition, spotting, and spatio-temporal localization in soccer — current trends and research perspec- tives.ACM Trans. Intell. Syst. Technol., 17(2):1–37, 2026. 3

    2026

  77. [77]

    Jo ˜ao V . B. Soares and Avijit Shah. Action spotting using dense detection anchors revisited: Submission to the Soccer- Net challenge 2022.arXiv, abs/2206.07846, 2022

    work page arXiv 2022

  78. [78]

    Jo ˜ao V . B. Soares, Avijit Shah, and Topojoy Biswas. Tem- porally precise action spotting in soccer videos using dense detection anchors. InIEEE Int. Conf. Image Process. (ICIP), pages 2796–2800, Bordeaux, France, 2022. IEEE. 3

    2022

  79. [79]

    Body part-based representation learning for occluded person re-identification

    Vladimir Somers, Christophe De Vleeschouwer, and Alexandre Alahi. Body part-based representation learning for occluded person re-identification. InIEEE/CVF Win- ter Conf. Appl. Comput. Vis. (WACV), pages 1613–1623, Waikoloa, HI, USA, 2023. IEEE. 3

    2023

  80. [80]

    Mansourian, Xin Zhou, Shohreh Kasaei, Bernard Ghanem, Alexandre Alahi, Marc Van Droogenbroeck, and Christophe De Vleeschouwer

    Vladimir Somers, Victor Joos, Anthony Cioppa, Silvio Gi- ancola, Seyed Abolfazl Ghasemzadeh, Floriane Magera, Baptiste Standaert, Amir M. Mansourian, Xin Zhou, Shohreh Kasaei, Bernard Ghanem, Alexandre Alahi, Marc Van Droogenbroeck, and Christophe De Vleeschouwer. Soc- cerNet game state reconstruction: End-to-end athlete track- ing and identification on a...

    2024

Showing first 80 references.