EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

KSL-POSE: A Real-Time 2D Human Pose Estimation Method Based on Modified YOLOv8-Pose Framework

Tianyi Lu iD, Ke Cheng, Xuecheng Hua iD, Suning Qin iD

Sensors · 26.09.2024 · 24 (19) · 6249

DOI: 10.3390/s24196249 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Sensors, online erschienen am 26.09.2024. Verfasst von Tianyi Lu und 3 weiteren Autor:innen. Beteiligte Einrichtung: School of Computer, Jiangsu University of Science and Technology, Zhenjiang 212100, China. Hauptthemen laut MEDLINE: Posture, Algorithms. Die Arbeit wurde bisher 16-mal zitiert (Crossref) und nennt 36 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: National Science Foundation Program of China (NSFC), Postgraduate Research and Practice Innovation Program of Jiangsu Province, International Science and technology cooperationplan project of Zhenjiang.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Two-dimensional human pose estimation aims to equip computers with the ability to accurately recognize human keypoints and comprehend their spatial contexts within media content. However, the accuracy of real-time human pose estimation diminishes when processing images with occluded body parts or overlapped individuals. To address these issues, we propose a method based on the YOLO framework. We integrate the convolutional concepts of Kolmogorov–Arnold Networks (KANs) through introducing non-linear activation functions to enhance the feature extraction capabilities of the convolutional kernels. Moreover, to improve the detection of small target keypoints, we integrate the cross-stage partial (CSP) approach and utilize the small object enhance pyramid (SOEP) module for feature integration. We also innovatively incorporate a layered shared convolution with batch normalization detection head (LSCB), consisting of multiple shared convolutional layers and batch normalization layers, to enable cross-stage feature fusion and address the low utilization of model parameters. Given the structure and purpose of the proposed model, we name it KSL-POSE. Compared to the baseline model YOLOv8l-POSE, KSL-POSE achieves significant improvements, increasing the average detection accuracy by 1.5% on the public MS COCO 2017 data set. Furthermore, the model also demonstrates competitive performance on the CrowdPOSE data set, thus validating its generalization ability.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

PostureAlgorithmsHumansImage Processing, Computer-AssistedNeural Networks, Computer

Stichwörter der Autor:innen

Kolmogorov–Arnold networksYOLOv8-posehuman pose estimationmulti-scale feature fusion

Autor:innen und Institutionen

  • School of Computer, Jiangsu University of Science and Technology, Zhenjiang 212100, China.

ORCID-Profile: Tianyi Lu, Xuecheng Hua, Suning Qin

Förderung

  • National Science Foundation Program of China (NSFC)
  • Postgraduate Research and Practice Innovation Program of Jiangsu Province
  • International Science and technology cooperationplan project of Zhenjiang

Zitieren

APA

Lu, T., Cheng, K., Hua, X., & Qin, S. (2024). KSL-POSE: A Real-Time 2D Human Pose Estimation Method Based on Modified YOLOv8-Pose Framework. Sensors, 24(19), 6249. https://doi.org/10.3390/s24196249

Vancouver

Lu T, Cheng K, Hua X, Qin S. KSL-POSE: A Real-Time 2D Human Pose Estimation Method Based on Modified YOLOv8-Pose Framework. Sensors. 2024;24(19):6249. doi:10.3390/s24196249

BibTeX
@article{Lu2024,
  author = {Lu, Tianyi and Cheng, Ke and Hua, Xuecheng and Qin, Suning},
  title = {KSL-POSE: A Real-Time 2D Human Pose Estimation Method Based on Modified YOLOv8-Pose Framework},
  journal = {Sensors},
  year = {2024},
  volume = {24},
  number = {19},
  pages = {6249},
  doi = {10.3390/s24196249}
}
Kostenlos starten →