EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Robust 2D Human Pose Estimation with Parallel Graph–Attention Modeling and Entropy-Aware Feature Decoding

Jiayuan Zhao iD, Dingyao Yu iD, Chunjia Han, Yingcheng Xu, Chunlei Shi

Entropy · 28.02.2026 · 28 (3) · 265

DOI: 10.3390/e28030265 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Entropy, online erschienen am 28.02.2026. Verfasst von Jiayuan Zhao und 4 weiteren Autor:innen. Beteiligte Einrichtung: School of Management, Harbin University of Commerce, Harbin 150028, China. Die Arbeit nennt 75 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: the central basic business research funding project, the China Postdoctoral Science Foundation under Grant, Central Basic Business Research Funding Project.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Robust 2D human pose estimation remains challenging due to occlusion and background interference, which introduce substantial uncertainty into visual representations. This paper proposes PMNet, a Parallel Modeling Network that integrates explicit graph-based structural modeling and implicit self-attention-based semantic modeling through parallel pathways to jointly capture local dependencies and global contextual relationships among keypoints. From an information-theoretic perspective, occlusion and clutter can be interpreted as sources of increased representational entropy, and PMNet addresses this issue by progressively reducing uncertainty through complementary structural reasoning and attention-based information selection. The framework incorporates a criss-cross attention module to suppress irrelevant features, an adaptive nonlinear fusion strategy to balance complementary information across parallel branches, and an error-compensated decoding method to sharpen heatmap distributions and refine keypoint localization while maintaining efficiency. Extensive experiments on the MPII and COCO benchmarks demonstrate that PMNet achieves state-of-the-art or comparable performance, attaining 92.42% PCKh@0.5 on MPII and 77.3% AP on COCO. Ablation studies and qualitative visualizations further confirm the effectiveness of each component, showing improved signal-to-noise ratios and more concentrated heatmap responses. Overall, PMNet provides a robust and efficient pose estimation framework with strong potential for real-world applications such as surveillance and autonomous systems.

Schlagwörter

Stichwörter der Autor:innen

attention mechanismentropy reductionexplicit modelinggraph neural networkhuman pose estimationimplicit modelingparallel structure

Autor:innen und Institutionen

  • School of Management, Harbin University of Commerce, Harbin 150028, China.
  • China Academy of Civil Aviation Science and Technology, Beijing 100028, China.
  • Business School, Birkbeck, University of London, Malet Street, Bloomsbury, London WC1E 7HX, UK.
  • China National Institute of Standardization, Beijing 100191, China.
  • School of Economics and Management, East University of Heilongjiang, Harbin 150066, China.

ORCID-Profile: Jiayuan Zhao, Dingyao Yu

Förderung

  • the central basic business research funding project
  • the China Postdoctoral Science Foundation under Grant
  • Central Basic Business Research Funding Project

Zitieren

APA

Zhao, J., Yu, D., Han, C., Xu, Y., & Shi, C. (2026). Robust 2D Human Pose Estimation with Parallel Graph–Attention Modeling and Entropy-Aware Feature Decoding. Entropy, 28(3), 265. https://doi.org/10.3390/e28030265

Vancouver

Zhao J, Yu D, Han C, Xu Y, Shi C. Robust 2D Human Pose Estimation with Parallel Graph–Attention Modeling and Entropy-Aware Feature Decoding. Entropy. 2026;28(3):265. doi:10.3390/e28030265

BibTeX
@article{Zhao2026,
  author = {Zhao, Jiayuan and Yu, Dingyao and Han, Chunjia and Xu, Yingcheng and Shi, Chunlei},
  title = {Robust 2D Human Pose Estimation with Parallel Graph–Attention Modeling and Entropy-Aware Feature Decoding},
  journal = {Entropy},
  year = {2026},
  volume = {28},
  number = {3},
  pages = {265},
  doi = {10.3390/e28030265}
}
Kostenlos starten →