IoT-enhanced multi-attention and lightweight feature integration for human pose estimation in motion training systems

Human pose estimation is widely used in intelligent sports training, rehabilitation assistance, and human–computer interaction, providing precise motion feedback and training guidance. However, existing methods suffer from keypoint localization errors and insufficient global coherence in complex bac...

Full description

Saved in:
Bibliographic Details
Main Authors: Junwen Chen, Jian Yang, Zhiqun Wang
Format: Article
Language:English
Published: Elsevier 2025-08-01
Series:Alexandria Engineering Journal
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S1110016825005708
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849229333980250112
author Junwen Chen
Jian Yang
Zhiqun Wang
author_facet Junwen Chen
Jian Yang
Zhiqun Wang
author_sort Junwen Chen
collection DOAJ
description Human pose estimation is widely used in intelligent sports training, rehabilitation assistance, and human–computer interaction, providing precise motion feedback and training guidance. However, existing methods suffer from keypoint localization errors and insufficient global coherence in complex backgrounds, occlusions, and high-dynamic motion scenarios. To address these challenges, this paper proposes a hybrid IoT-vision deep learning model for human pose estimation and motion training feedback. IoT-based motion sensors are integrated with vision-based keypoint detection to enhance pose estimation accuracy, particularly in occluded or high-speed movement scenarios. The model employs the LSFE stacked feature extraction module to enhance multi-scale feature adaptability, incorporates the LFAM local attention mechanism (SPAM + CARM) to improve key joint modeling, and introduces the GEAM global enhancement module to ensure keypoint stability and consistency. Additionally, an ECA-based lightweight channel attention mechanism reduces computational complexity while enhancing key feature responses. Experimental results show that the proposed model achieves Mean Accuracy of 0.946 and 0.949 on the LSP and MPII datasets, respectively, with PCK scores of 0.97 and 0.95. This model demonstrates significant improvements over existing methods in real-time performance and robustness, particularly in complex scenarios such as sports training, rehabilitation, and monitoring.
format Article
id doaj-art-cb629594e5124b059f2cefcc9d0e8d48
institution Kabale University
issn 1110-0168
language English
publishDate 2025-08-01
publisher Elsevier
record_format Article
series Alexandria Engineering Journal
spelling doaj-art-cb629594e5124b059f2cefcc9d0e8d482025-08-22T04:55:15ZengElsevierAlexandria Engineering Journal1110-01682025-08-0112728429510.1016/j.aej.2025.04.074IoT-enhanced multi-attention and lightweight feature integration for human pose estimation in motion training systemsJunwen Chen0Jian Yang1Zhiqun Wang2Jiangxi University of Finance and Economics, Nanchang, Jiangxi 330000, ChinaDepartment of Physical Education, Zhejiang University of Finance & Economics, Hangzhou, 310018, China; Corresponding author.School of Electronic Information, HuZhou College, HuZhou, 313000, ChinaHuman pose estimation is widely used in intelligent sports training, rehabilitation assistance, and human–computer interaction, providing precise motion feedback and training guidance. However, existing methods suffer from keypoint localization errors and insufficient global coherence in complex backgrounds, occlusions, and high-dynamic motion scenarios. To address these challenges, this paper proposes a hybrid IoT-vision deep learning model for human pose estimation and motion training feedback. IoT-based motion sensors are integrated with vision-based keypoint detection to enhance pose estimation accuracy, particularly in occluded or high-speed movement scenarios. The model employs the LSFE stacked feature extraction module to enhance multi-scale feature adaptability, incorporates the LFAM local attention mechanism (SPAM + CARM) to improve key joint modeling, and introduces the GEAM global enhancement module to ensure keypoint stability and consistency. Additionally, an ECA-based lightweight channel attention mechanism reduces computational complexity while enhancing key feature responses. Experimental results show that the proposed model achieves Mean Accuracy of 0.946 and 0.949 on the LSP and MPII datasets, respectively, with PCK scores of 0.97 and 0.95. This model demonstrates significant improvements over existing methods in real-time performance and robustness, particularly in complex scenarios such as sports training, rehabilitation, and monitoring.http://www.sciencedirect.com/science/article/pii/S1110016825005708Human pose estimationMotion training feedbackLightweight attention mechanismMulti-scale feature extractionKeypoint detectionIoT-based sensor fusion
spellingShingle Junwen Chen
Jian Yang
Zhiqun Wang
IoT-enhanced multi-attention and lightweight feature integration for human pose estimation in motion training systems
Alexandria Engineering Journal
Human pose estimation
Motion training feedback
Lightweight attention mechanism
Multi-scale feature extraction
Keypoint detection
IoT-based sensor fusion
title IoT-enhanced multi-attention and lightweight feature integration for human pose estimation in motion training systems
title_full IoT-enhanced multi-attention and lightweight feature integration for human pose estimation in motion training systems
title_fullStr IoT-enhanced multi-attention and lightweight feature integration for human pose estimation in motion training systems
title_full_unstemmed IoT-enhanced multi-attention and lightweight feature integration for human pose estimation in motion training systems
title_short IoT-enhanced multi-attention and lightweight feature integration for human pose estimation in motion training systems
title_sort iot enhanced multi attention and lightweight feature integration for human pose estimation in motion training systems
topic Human pose estimation
Motion training feedback
Lightweight attention mechanism
Multi-scale feature extraction
Keypoint detection
IoT-based sensor fusion
url http://www.sciencedirect.com/science/article/pii/S1110016825005708
work_keys_str_mv AT junwenchen iotenhancedmultiattentionandlightweightfeatureintegrationforhumanposeestimationinmotiontrainingsystems
AT jianyang iotenhancedmultiattentionandlightweightfeatureintegrationforhumanposeestimationinmotiontrainingsystems
AT zhiqunwang iotenhancedmultiattentionandlightweightfeatureintegrationforhumanposeestimationinmotiontrainingsystems