SCite‐HRNet: A self‐calibrating efficient network for pose estimation

Abstract In order to tackle the challenge of capturing long‐range spatial dependencies among joints, a novel self‐calibrated lightweight high‐resolution network (SCite‐HRNet), which is grounded on the lightweight high‐resolution network is introduced. A self‐calibrated segmentation convolution is fi...

Full description

Saved in:
Bibliographic Details
Main Authors: Nan Xiang, Xingdi Rao, Wenjing Yang, Jin Chen, Lifang Zhu
Format: Article
Language:English
Published: Wiley 2024-12-01
Series:Electronics Letters
Subjects:
Online Access:https://doi.org/10.1049/ell2.70106
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850108607197085696
author Nan Xiang
Xingdi Rao
Wenjing Yang
Jin Chen
Lifang Zhu
author_facet Nan Xiang
Xingdi Rao
Wenjing Yang
Jin Chen
Lifang Zhu
author_sort Nan Xiang
collection DOAJ
description Abstract In order to tackle the challenge of capturing long‐range spatial dependencies among joints, a novel self‐calibrated lightweight high‐resolution network (SCite‐HRNet), which is grounded on the lightweight high‐resolution network is introduced. A self‐calibrated segmentation convolution is first designed to extract and amalgamate contextual information across various scales, thereby addressing the issue of excessive computation engendered by stacked convolution kernels in conventional convolution methods. Next, a multi‐scale channel attention mechanism designed to extract structural information while filtering out unnecessary channel details is introduced. Ultimately, these two methodologies are incorporated into a multi‐scale information aggregation module and embed this module into the high‐resolution network. This allows the network to maintain exceptionally efficient computation while effectively managing information across various scales. Empirical results indicate that SCite‐HRNet achieves remarkable performance on both the COCO dataset and the challenging average precision (AP)‐10K dataset.
format Article
id doaj-art-4bf5be0719cb423e87a1c4c59e4c0239
institution OA Journals
issn 0013-5194
1350-911X
language English
publishDate 2024-12-01
publisher Wiley
record_format Article
series Electronics Letters
spelling doaj-art-4bf5be0719cb423e87a1c4c59e4c02392025-08-20T02:38:18ZengWileyElectronics Letters0013-51941350-911X2024-12-016023n/an/a10.1049/ell2.70106SCite‐HRNet: A self‐calibrating efficient network for pose estimationNan Xiang0Xingdi Rao1Wenjing Yang2Jin Chen3Lifang Zhu4Liangjiang International College Chongqing University of Technology Chongqing ChinaLiangjiang Artificial Intelligence College Chongqing University of Technology Chongqing ChinaComputer Science and Engineering College Chongqing University of Technology Chongqing ChinaChongqing Jialing Special Equipment Co., Ltd. Chongqing ChinaChongqing Jialing Special Equipment Co., Ltd. Chongqing ChinaAbstract In order to tackle the challenge of capturing long‐range spatial dependencies among joints, a novel self‐calibrated lightweight high‐resolution network (SCite‐HRNet), which is grounded on the lightweight high‐resolution network is introduced. A self‐calibrated segmentation convolution is first designed to extract and amalgamate contextual information across various scales, thereby addressing the issue of excessive computation engendered by stacked convolution kernels in conventional convolution methods. Next, a multi‐scale channel attention mechanism designed to extract structural information while filtering out unnecessary channel details is introduced. Ultimately, these two methodologies are incorporated into a multi‐scale information aggregation module and embed this module into the high‐resolution network. This allows the network to maintain exceptionally efficient computation while effectively managing information across various scales. Empirical results indicate that SCite‐HRNet achieves remarkable performance on both the COCO dataset and the challenging average precision (AP)‐10K dataset.https://doi.org/10.1049/ell2.70106artificial intelligencecomputer visionconvolutionimage classification
spellingShingle Nan Xiang
Xingdi Rao
Wenjing Yang
Jin Chen
Lifang Zhu
SCite‐HRNet: A self‐calibrating efficient network for pose estimation
Electronics Letters
artificial intelligence
computer vision
convolution
image classification
title SCite‐HRNet: A self‐calibrating efficient network for pose estimation
title_full SCite‐HRNet: A self‐calibrating efficient network for pose estimation
title_fullStr SCite‐HRNet: A self‐calibrating efficient network for pose estimation
title_full_unstemmed SCite‐HRNet: A self‐calibrating efficient network for pose estimation
title_short SCite‐HRNet: A self‐calibrating efficient network for pose estimation
title_sort scite hrnet a self calibrating efficient network for pose estimation
topic artificial intelligence
computer vision
convolution
image classification
url https://doi.org/10.1049/ell2.70106
work_keys_str_mv AT nanxiang scitehrnetaselfcalibratingefficientnetworkforposeestimation
AT xingdirao scitehrnetaselfcalibratingefficientnetworkforposeestimation
AT wenjingyang scitehrnetaselfcalibratingefficientnetworkforposeestimation
AT jinchen scitehrnetaselfcalibratingefficientnetworkforposeestimation
AT lifangzhu scitehrnetaselfcalibratingefficientnetworkforposeestimation