Multi-person 2D human pose estimation a benchmark for real-time applications
Engineering review · 04.08.2026 · 46 (2)
DOI: 10.30765/er.3423 ↗Kurz gefasst
Fachartikel in der Zeitschrift Engineering review, online erschienen am 04.08.2026. Verfasst von Tomislav Prusina und 3 weiteren Autor:innen.
Abstract (Originalsprache)
Human pose estimation (HPE) is a critical component of computer vision, enabling real-time applications across various fields. This study benchmarks six state-of-the-art frameworks: OpenPose, YOLO, RTMO, RTMPose, Sapiens, and MoveNet in the context of 2D multi-person pose estimation in videos for real-time applications. Using the Panoptic dataset and standardized metrics, including Object Keypoint Similarity (OKS), Average Precision (AP), and Average Recall (AR), we evaluate their accuracy, speed, and hardware efficiency under realistic conditions such as occlusion and crowded scenes. Our analysis highlights the strengths and limitations of each framework, providing valuable insights to help practitioners select and deploy reliable HPE solutions in real-world applications.
Zitieren
APA
Prusina, T., Benić, J., Ševerdija, D., & Matijević, D. (2026). Multi-person 2D human pose estimation a benchmark for real-time applications. Engineering review, 46(2). https://doi.org/10.30765/er.3423
Vancouver
Prusina T, Benić J, Ševerdija D, Matijević D. Multi-person 2D human pose estimation a benchmark for real-time applications. Engineering review. 2026;46(2). doi:10.30765/er.3423
BibTeX
@article{Prusina2026,
author = {Prusina, Tomislav and Benić, Juraj and Ševerdija, Domagoj and Matijević, Domagoj},
title = {Multi-person 2D human pose estimation a benchmark for real-time applications},
journal = {Engineering review},
year = {2026},
volume = {46},
number = {2},
doi = {10.30765/er.3423}
}
EUVIMED