Advanced search
2 files | 12.22 MB Add to list

One-shot team recognition and 3D pose estimation of cyclists for augmented reality visualization

Author
Organization
Abstract
Advanced computer vision and machine learning technologies transform how we experience sports events. This research focuses on enhancing the viewing experience of cycling races by automatically identifying teams from helicopter footage. It employs a multi-stage pipeline that tackles challenges such as rapid motion and similar team uniforms. Initially, cyclists are detected and tracked. Team recognition is then performed using a one-shot learning approach based on Siamese neural networks, achieving a classification accuracy of 85% on a test set composed of previously unseen teams. This method reduces the need for extensive labeling. Additionally, temporal post-processing techniques, such as applying a moving average to confidence scores, further enhance classification performance. These methods ensure reliable identification of teams and track their presence throughout the race footage. Furthermore, we integrate 3D pose estimation to generate augmented reality (AR) overlays that display rider-specific information, such as names and speeds, enhancing the broadcast’s informational value. The combination of advanced computer vision and AR showcases new possibilities for improving live sports broadcasts, particularly in challenging environments like road cycling.
Keywords
Computer vision, Object detection, One-shot classification, 3D Pose Estimation, Augmented Reality, Sports Broadcasting

Downloads

  • (...).pdf
    • full text (Published version)
    • |
    • UGent only
    • |
    • PDF
    • |
    • 1.31 MB
  • (...).pdf
    • full text (Accepted manuscript)
    • |
    • UGent only (changes to open access on 2027-09-27)
    • |
    • PDF
    • |
    • 10.90 MB

Citation

Please use this url to cite or link to this publication:

MLA
Clinckemaillie, Winter, et al. “One-Shot Team Recognition and 3D Pose Estimation of Cyclists for Augmented Reality Visualization.” SPORTS ANALYTICS, ISACE 2025, edited by Jin-song Dong et al., vol. 15925, Springer, 2025, pp. 36–52, doi:10.1007/978-3-032-06167-6_3.
APA
Clinckemaillie, W., Vanhaeverbeke, J., Slembrouck, M., & Verstockt, S. (2025). One-shot team recognition and 3D pose estimation of cyclists for augmented reality visualization. In J. Dong, J. Sun, X. Xie, & K. Jiang (Eds.), SPORTS ANALYTICS, ISACE 2025 (Vol. 15925, pp. 36–52). https://doi.org/10.1007/978-3-032-06167-6_3
Chicago author-date
Clinckemaillie, Winter, Jelle Vanhaeverbeke, Maarten Slembrouck, and Steven Verstockt. 2025. “One-Shot Team Recognition and 3D Pose Estimation of Cyclists for Augmented Reality Visualization.” In SPORTS ANALYTICS, ISACE 2025, edited by Jin-song Dong, Jing Sun, Xiaofei Xie, and Kan Jiang, 15925:36–52. Springer. https://doi.org/10.1007/978-3-032-06167-6_3.
Chicago author-date (all authors)
Clinckemaillie, Winter, Jelle Vanhaeverbeke, Maarten Slembrouck, and Steven Verstockt. 2025. “One-Shot Team Recognition and 3D Pose Estimation of Cyclists for Augmented Reality Visualization.” In SPORTS ANALYTICS, ISACE 2025, ed by. Jin-song Dong, Jing Sun, Xiaofei Xie, and Kan Jiang, 15925:36–52. Springer. doi:10.1007/978-3-032-06167-6_3.
Vancouver
1.
Clinckemaillie W, Vanhaeverbeke J, Slembrouck M, Verstockt S. One-shot team recognition and 3D pose estimation of cyclists for augmented reality visualization. In: Dong J, Sun J, Xie X, Jiang K, editors. SPORTS ANALYTICS, ISACE 2025. Springer; 2025. p. 36–52.
IEEE
[1]
W. Clinckemaillie, J. Vanhaeverbeke, M. Slembrouck, and S. Verstockt, “One-shot team recognition and 3D pose estimation of cyclists for augmented reality visualization,” in SPORTS ANALYTICS, ISACE 2025, Shanghai, China, 2025, vol. 15925, pp. 36–52.
@inproceedings{01K7EJ6CSB2Q06J86PD739FSVN,
  abstract     = {{Advanced computer vision and machine learning technologies transform how we experience sports events. This research focuses on enhancing the viewing experience of cycling races by automatically identifying teams from helicopter footage. It employs a multi-stage pipeline that tackles challenges such as rapid motion and similar team uniforms. Initially, cyclists are detected and tracked. Team recognition is then performed using a one-shot learning approach based on Siamese neural networks, achieving a classification accuracy of 85% on a test set composed of previously unseen teams. This method reduces the need for extensive labeling. Additionally, temporal post-processing techniques, such as applying a moving average to confidence scores, further enhance classification performance. These methods ensure reliable identification of teams and track their presence throughout the race footage. Furthermore, we integrate 3D pose estimation to generate augmented reality (AR) overlays that display rider-specific information, such as names and speeds, enhancing the broadcast’s informational value. The combination of advanced computer vision and AR showcases new possibilities for improving live sports broadcasts, particularly in challenging environments like road cycling.}},
  author       = {{Clinckemaillie, Winter and Vanhaeverbeke, Jelle and Slembrouck, Maarten and Verstockt, Steven}},
  booktitle    = {{SPORTS ANALYTICS, ISACE 2025}},
  editor       = {{Dong, Jin-song and Sun, Jing and Xie, Xiaofei and Jiang, Kan}},
  isbn         = {{9783032061669}},
  issn         = {{0302-9743}},
  keywords     = {{Computer vision,Object detection,One-shot classification,3D Pose Estimation,Augmented Reality,Sports Broadcasting}},
  language     = {{eng}},
  location     = {{Shanghai, China}},
  pages        = {{36--52}},
  publisher    = {{Springer}},
  title        = {{One-shot team recognition and 3D pose estimation of cyclists for augmented reality visualization}},
  url          = {{http://doi.org/10.1007/978-3-032-06167-6_3}},
  volume       = {{15925}},
  year         = {{2025}},
}

Altmetric
View in Altmetric
Web of Science
Times cited: