{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,25]],"date-time":"2026-08-25T12:23:44Z","timestamp":1787660624727,"version":"build-2736575974"},"reference-count":35,"publisher":"Public Library of Science (PLoS)","issue":"9","license":[{"start":{"date-parts":[[2025,9,2]],"date-time":"2025-09-02T00:00:00Z","timestamp":1756771200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100011844","name":"Science and Technology Department, Heilongjiang Province","doi-asserted-by":"publisher","award":["GZ20220131"],"award-info":[{"award-number":["GZ20220131"]}],"id":[{"id":"10.13039\/501100011844","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS One"],"abstract":"<jats:p>Scale variation is a challenge in human pose estimation. The scale variations of human body are related to the accuracy and robustness of posture estimation. For example, the prediction accuracy of smaller joints (such as ankles and wrists) is less than that of larger joints (such as head and shoulders). To address the impact of scale variations across parts of the human body on the positioning of key points. In this paper, we propose a Detail Enhanced High-Resolution Network (DE-HRNet), which can efficiently extract local detail features and mitigate the impact of scale variations for human pose estimation. First, we propose a Detail Enhancement Module (DEM) to relearn the lost low-level detailed features and enhance the model\u2019s ability to capture delicate local features, which is crucial for improving the accuracy of scale-varying keypoints. Second, we introduce an ultra-lightweight dynamic sampler - dySample, which is used to replace nearest up-sampling. It aims to reduce the loss of detail information from low-resolution features during up-sampling, while simultaneously preserving finer local representations for high resolution, it can be beneficial in improving the robustness of the model in dealing with scale-varying keypoints. On the COCO test-dev2017 and MPII valid datasets, our method achieved 75.6 AP and 90.7 PCKh@0.5, respectively, compared to High-Resolution Network (HRNet), it improved by 0.7 and 0.4 points. In comparison with the other works, the proposed method has performed well in the scale variation.<\/jats:p>","DOI":"10.1371\/journal.pone.0325540","type":"journal-article","created":{"date-parts":[[2025,9,2]],"date-time":"2025-09-02T17:49:24Z","timestamp":1756835364000},"page":"e0325540","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":5,"title":["DE-HRNet: Detail enhanced high-resolution network for human pose estimation"],"prefix":"10.1371","volume":"20","author":[{"ORCID":"https:\/\/orcid.org\/0009-0003-2979-7877","authenticated-orcid":true,"given":"Yuxuan","family":"Liu","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-4500-1002","authenticated-orcid":true,"given":"Guohui","family":"Zhou","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Wei","family":"He","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hailong","family":"Zhu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yanling","family":"Cui","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2025,9,2]]},"reference":[{"key":"pone.0325540.ref001","first-page":"3578","year":"2018"},{"key":"pone.0325540.ref002","first-page":"5992","article-title":"Scale-aware modulation meet transformer.","year":"2023"},{"issue":"4","key":"pone.0325540.ref003","first-page":"5296","article-title":"ZoomNAS: searching for whole-body human pose estimation in the wild","volume":"45","author":"L Xu","year":"2023","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"pone.0325540.ref004","first-page":"5385","article-title":"HigherHRNet: Scale-aware representation learning for bottom-up human pose estimation.","year":"2020"},{"key":"pone.0325540.ref005","first-page":"5686","article-title":"Deep High-Resolution Representation Learning for Human Pose Estimation.","year":"2019"},{"key":"pone.0325540.ref006","unstructured":"Yuan Y, Fu R, Huang L. HRFormer: high-resolution transformer for dense prediction. In: Proceedings of the 35th International Conference on Neural Information Processing Systems. 2021. p. 7281\u201393."},{"key":"pone.0325540.ref007","first-page":"13259","article-title":"Rethinking the heatmap regression for bottom-up human pose estimation.","year":"2021"},{"key":"pone.0325540.ref008","doi-asserted-by":"crossref","first-page":"455","DOI":"10.1007\/978-3-030-58580-8_27","article-title":"Learning delicate local representations for multi-person pose estimation.","author":"Y Cai","year":"2020","journal-title":"Lecture Notes in Computer Science"},{"issue":"10","key":"pone.0325540.ref009","doi-asserted-by":"crossref","first-page":"3349","DOI":"10.1109\/TPAMI.2020.2983686","article-title":"Deep high-resolution representation learning for visual recognition","volume":"43","author":"J Wang","year":"2021","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"pone.0325540.ref010","article-title":"Efficient object localization using Convolutional Networks.","year":"2015"},{"key":"pone.0325540.ref011","first-page":"10435","article-title":"Lite-HRNet: A lightweight high-resolution network.","year":"2021"},{"key":"pone.0325540.ref012","first-page":"7103","article-title":"Cascaded pyramid network for multi-person pose estimation.","year":"2018"},{"key":"pone.0325540.ref013","first-page":"1653","article-title":"DeepPose: Human pose estimation via deep neural networks.","year":"2014"},{"key":"pone.0325540.ref014","first-page":"564","article-title":"Multi-scale supervised network for human pose estimation.","year":"2018"},{"key":"pone.0325540.ref015","doi-asserted-by":"crossref","first-page":"483","DOI":"10.1007\/978-3-319-46484-8_29","article-title":"Stacked hourglass networks for human pose estimation.","author":"A Newell","year":"2016","journal-title":"Lecture Notes in Computer Science"},{"key":"pone.0325540.ref016","doi-asserted-by":"crossref","unstructured":"Lin TY, Doll\u00e1r P, Girshick R. Feature pyramid networks for object detection. In: Proceedings of the IEEE conference on computer vision and pattern recognition. 2017. p. 2117\u201325.","DOI":"10.1109\/CVPR.2017.106"},{"key":"pone.0325540.ref017","doi-asserted-by":"crossref","unstructured":"Liu W, Lu H, Fu H. Learning to upsample by learning to sample. In: Proceedings of the IEEE\/CVF International Conference on Computer Vision. 2023. p. 6027\u201337.","DOI":"10.1109\/ICCV51070.2023.00554"},{"issue":"12","key":"pone.0325540.ref018","doi-asserted-by":"crossref","first-page":"10763","DOI":"10.1109\/TPAMI.2024.3449959","article-title":"Frequency-aware feature fusion for dense image prediction","volume":"46","author":"L Chen","year":"2024","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"pone.0325540.ref019","first-page":"6841","article-title":"Learning affinity-aware upsampling for deep image matting.","author":"Y Dai","year":"2021","journal-title":"Proceedings of IEEE Conference on Computer Vision and Pattern Recognition"},{"key":"pone.0325540.ref020","unstructured":"Lu H, Liu W, Ye Z, Fu H, Liu Y, Cao Z. Sapa: Similarity-aware point affiliation for feature upsampling. In: Proceedings of Advances in Neural Information Processing Systems. 2022. p. 1\u201312."},{"key":"pone.0325540.ref021","article-title":"Squeeze-and-excitation networks.","year":"2018"},{"key":"pone.0325540.ref022","doi-asserted-by":"crossref","first-page":"3","DOI":"10.1007\/978-3-030-01234-2_1","article-title":"CBAM: Convolutional block attention module.","author":"S Woo","year":"2018","journal-title":"Lecture Notes in Computer Science"},{"key":"pone.0325540.ref023","doi-asserted-by":"crossref","unstructured":"Wang Q, Wu B, Zhu P. ECA-Net: Efficient channel attention for deep convolutional neural networks. In: Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition, 2020. p. 11534\u201342.","DOI":"10.1109\/CVPR42600.2020.01155"},{"key":"pone.0325540.ref024","doi-asserted-by":"crossref","unstructured":"Xiao C, Yang W, Ouyang W, Ma C, Yuille AL, Wang X. Multi-context attention for human pose estimation. In: CVPR. 2017. p. 5669\u201378.","DOI":"10.1109\/CVPR.2017.601"},{"key":"pone.0325540.ref025","article-title":"Improving neural networks by preventing co-adaptation of feature detectors.","author":"GE Hinton","year":"2012"},{"key":"pone.0325540.ref026","doi-asserted-by":"crossref","first-page":"740","DOI":"10.1007\/978-3-319-10602-1_48","article-title":"Microsoft COCO: common objects in context.","author":"T-Y Lin","year":"2014","journal-title":"Lecture Notes in Computer Science"},{"key":"pone.0325540.ref027","article-title":"2D human pose estimation: new benchmark and state of the art analysis.","year":"2014"},{"key":"pone.0325540.ref028","unstructured":"Kingma DP, Ba J. Adam: A method for stochastic optimization. CoRR. 2014;abs\/1412.6980."},{"issue":"6","key":"pone.0325540.ref029","doi-asserted-by":"crossref","first-page":"1137","DOI":"10.1109\/TPAMI.2016.2577031","article-title":"Faster R-CNN: towards real-time object detection with region proposal networks","volume":"39","author":"S Ren","year":"2017","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"key":"pone.0325540.ref030","doi-asserted-by":"crossref","unstructured":"Zhang F, Zhu X, Dai H. Distribution-aware coordinate representation for human pose estimation. In: Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition. 2020. p. 7093\u2013102.","DOI":"10.1109\/CVPR42600.2020.00712"},{"key":"pone.0325540.ref031","doi-asserted-by":"crossref","first-page":"472","DOI":"10.1007\/978-3-030-01231-1_29","article-title":"Simple baselines for human pose estimation and tracking.","author":"B Xiao","year":"2018","journal-title":"Lecture Notes in Computer Science"},{"key":"pone.0325540.ref032","article-title":"A coarse-fine network for keypoint localization.","year":"2017"},{"key":"pone.0325540.ref033","article-title":"Rethinking on multi-stage networks for human pose estimation.","author":"W Li","year":"2019"},{"key":"pone.0325540.ref034","doi-asserted-by":"crossref","unstructured":"Li Y, Zhang S, Wang Z. Tokenpose: Learning keypoint tokens for human pose estimation. In: Proceedings of the IEEE\/CVF International Conference on Computer Vision. 2021. p. 11313\u201322.","DOI":"10.1109\/ICCV48922.2021.01112"},{"key":"pone.0325540.ref035","doi-asserted-by":"crossref","unstructured":"Huang J, Zhu Z, Guo F. The devil is in the details: delving into unbiased data processing for human pose estimation. In: Proceedings of the IEEE\/CVF conference on computer vision and pattern recognition. 2020. p. 5700\u20139.","DOI":"10.1109\/CVPR42600.2020.00574"}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0325540","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,9,2]],"date-time":"2025-09-02T17:49:32Z","timestamp":1756835372000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0325540"}},"subtitle":[],"editor":[{"given":"Yongjie","family":"Li","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2025,9,2]]},"references-count":35,"journal-issue":{"issue":"9","published-online":{"date-parts":[[2025,9,2]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0325540","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,9,2]]}}}