Toward Automatically Labeling Situations in Soccer

Research output: Journal contributionsJournal articlesResearchpeer-review

Authors

We study the automatic annotation of situations in soccer games. At first sight, this translates nicely into a standard supervised learning problem. However, in a fully supervised setting, predictive accuracies are supposed to correlate positively with the amount of labeled situations: more labeled training data simply promise better performance. Unfortunately, non-trivially annotated situations in soccer games are scarce, expensive and almost always require human experts; a fully supervised approach appears infeasible. Hence, we split the problem into two parts and learn (i) a meaningful feature representation using variational autoencoders on unlabeled data at large scales and (ii) a large-margin classifier acting in this feature space but utilize only a few (manually) annotated examples of the situation of interest. We propose four different architectures of the variational autoencoder and empirically study the detection of corner kicks, crosses and counterattacks. We observe high predictive accuracies above 90% AUC irrespectively of the task.

Original languageEnglish
Article number725431
JournalFrontiers in Sports and Active Living
Volume3
Number of pages15
DOIs
Publication statusPublished - 03.11.2021

Bibliographical note

This publication was funded by the Open Access Publication Fund of Leuphana University Lüneburg.

We would like to thank the German Football Association (DFB) for providing the data for this study.

Documents

DOI

Recently viewed

Publications

  1. Studying properties of water data using manifold-aware anomaly detectors
  2. To use or not to use learning data
  3. Influence of data clouds fusion from 3D real-time vision system on robotic group dead reckoning in unknown terrain
  4. Endemic predators, invasive prey and native diversity
  5. Decision making in the context of environmental risks
  6. Performability analysis of an unreliable M/M/1-type queue
  7. Effects of tree diversity on canopy space occupation vary with tree size and canopy space definition in a mature broad-leaved forest
  8. Who’s afraid of the senses? Organization, management and the return of the sensorium
  9. Assessing the Bonding Interface Characteristics and Mechanical Properties of Bobbin Tool Friction Stir Welded Dissimilar Aluminum Alloy Joints
  10. Optimization of transport flow on two paths with respect to the passengers time costs
  11. Transformationsnarrative
  12. Structure and Organization of Product Development Projects
  13. Atomkraft international
  14. Consumer Online Search Behavior
  15. Assessing teachers' educational knowledge
  16. On the measuring accuracy of the “Vehrs-Hebel”, a scaling apparatus for nonverbal real-time assessment of perceived quantity
  17. A control strategy for electromagnetic near and far field calculation
  18. Art and Culture as an Urban Development Tool
  19. A(l)gora: the Mindscape
  20. “Processed Food on the Urban Data Highway. Food Delivery Services as In_Visible Infrastructure in the Production of Urbanity”
  21. Careless product use in access-based services
  22. Quantum Chemical Calculation and Evaluation of Partition Coefficients for Classical and Emerging Environmentally Relevant Organic Compounds