{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,28]],"date-time":"2026-08-28T18:19:13Z","timestamp":1787941153050,"version":"build-2784847793"},"reference-count":36,"publisher":"Public Library of Science (PLoS)","issue":"8","license":[{"start":{"date-parts":[[2026,8,28]],"date-time":"2026-08-28T00:00:00Z","timestamp":1787875200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/100017724","name":"Henan Provincial Department of Transportation","doi-asserted-by":"publisher","award":["2023-2-1"],"award-info":[{"award-number":["2023-2-1"]}],"id":[{"id":"10.13039\/100017724","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100017700","name":"Henan Provincial Science and Technology Research Project","doi-asserted-by":"publisher","award":["SKXJCZX-2026-29B"],"award-info":[{"award-number":["SKXJCZX-2026-29B"]}],"id":[{"id":"10.13039\/501100017700","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100017700","name":"Henan Provincial Science and Technology Research Project","doi-asserted-by":"publisher","award":["262102521014"],"award-info":[{"award-number":["262102521014"]}],"id":[{"id":"10.13039\/501100017700","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["www.plosone.org"],"crossmark-restriction":false},"short-container-title":["PLoS One"],"abstract":"<jats:p>To address the challenges of false detection, insufficient feature representation, and limited real-time performance in image-based traffic police command gesture detection under different lighting conditions, this paper proposes a detection model named FDMB-YOLOv11 based on the YOLOv11n architecture. Firstly, the detection head was enhanced by introducing Frequency Adaptive Dilated Convolution (FADC), while the BiFormer attention mechanism was incorporated into the C2PSA module. Secondly, MobileNetV4 was adopted as the backbone network to reduce computational complexity. Finally, standard convolutions were replaced with Deformable Convolution to improve feature representation capability. Experimental results show that, based on the improvements described above, the precision, recall, and mean average precision (mAP@0.5) of the FDMB-YOLOv11 model under normal lighting conditions reach 97.91%, 93.32%, and 97.17% respectively, which are 23.1%, 13.63%, and 16.74% higher than those of the YOLOv11n model. The model\u2019s parameter count is reduced from 2.58 million to 1.93 million, achieving lightweight optimization. Although the frame rate (FPS) is reduced to 65.789 fps, the model still achieves real-time inference performance and demonstrates potential for deployment on edge devices. The model outperforms target detection algorithms like SSD, Faster R-CNN, RTDETR, YOLOv12, and YOLOv13 by achieving the highest mAP@0.5. It also addresses the misclassification problem seen in the YOLOv11n model regarding left-turn and right-turn gestures while maintaining an optimal balance between accuracy and parameter count.<\/jats:p>","DOI":"10.1371\/journal.pone.0357212","type":"journal-article","created":{"date-parts":[[2026,8,28]],"date-time":"2026-08-28T17:43:16Z","timestamp":1787938996000},"page":"e0357212","update-policy":"https:\/\/doi.org\/10.1371\/journal.pone.corrections_policy","source":"Crossref","is-referenced-by-count":0,"title":["FDMB-YOLOv11: Traffic police command gesture recognition method under different lighting conditions"],"prefix":"10.1371","volume":"21","author":[{"given":"Xiaoyu","family":"Zhang","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-7238-4109","authenticated-orcid":true,"given":"Baohua","family":"Guo","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1236-5328","authenticated-orcid":true,"given":"Sigama","family":"Anthony","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"David","family":"Bassir","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yuandi","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"340","published-online":{"date-parts":[[2026,8,28]]},"reference":[{"key":"pone.0357212.ref001","doi-asserted-by":"crossref","first-page":"129042","DOI":"10.1016\/j.neucom.2024.129042","article-title":"Low light recognition of traffic police gestures based on lightweight extraction of skeleton features","volume":"617","author":"M Chang","year":"2025","journal-title":"Neurocomputing"},{"key":"pone.0357212.ref002","doi-asserted-by":"crossref","first-page":"108282","DOI":"10.1016\/j.aap.2025.108282","article-title":"A new framework for modelling traffic conflict interval time based on correlated random parameter duration model with heterogeneity in means and variances","volume":"224","author":"G Li","year":"2026","journal-title":"Accid Anal Prev"},{"key":"pone.0357212.ref003","doi-asserted-by":"crossref","first-page":"107632","DOI":"10.1016\/j.scs.2026.107632","article-title":"How do water networks shape e-bike sharing near metro? A spatiotemporal analysis in Shaoxing, China","volume":"148","author":"G Li","year":"2026","journal-title":"Sustainable Cities and Society"},{"key":"pone.0357212.ref004","first-page":"500","volume-title":"Hand gesture recognition exploiting handcrafted features and lstm. In: International conference on image analysis and processing","author":"D Avola","year":"2023"},{"key":"pone.0357212.ref005","unstructured":"Choudhury ZH, Bhuyan HK, Brahma B, Kamila NK, et al. Adaptive Object Detection with ESRGAN-Enhanced Resolution & Faster R-CNN. 2025. https:\/\/doi.org\/10.48550\/arXiv.2506.11122"},{"key":"pone.0357212.ref006","doi-asserted-by":"crossref","first-page":"102210","DOI":"10.1016\/j.ecoinf.2023.102210","article-title":"Strawberry R-CNN: Recognition and counting model of strawberry based on improved faster R-CNN","volume":"77","author":"J Li","year":"2023","journal-title":"Ecological Informatics"},{"key":"pone.0357212.ref007","doi-asserted-by":"crossref","first-page":"4517","DOI":"10.1109\/ACCESS.2017.2684186","article-title":"Multimodal Gesture Recognition Using 3-D Convolution and Convolutional LSTM","volume":"5","author":"G Zhu","year":"2017","journal-title":"IEEE Access"},{"key":"pone.0357212.ref008","doi-asserted-by":"crossref","unstructured":"Tang Z, Tang J, Shang Y. A data-driven RetinaNet model for small object detection in aerial images. 2025. https:\/\/doi.org\/10.48550\/arXiv.2509.02928","DOI":"10.1109\/FMLDS67896.2025.00048"},{"key":"pone.0357212.ref009","doi-asserted-by":"crossref","first-page":"530","DOI":"10.1016\/j.aej.2025.03.077","article-title":"Assessing YOLO models for real-time object detection in urban environments for advanced driver-assistance systems (ADAS)","volume":"123","author":"R Ayachi","year":"2025","journal-title":"Alexandria Engineering Journal"},{"issue":"5","key":"pone.0357212.ref010","doi-asserted-by":"crossref","first-page":"1909","DOI":"10.18280\/ts.400511","article-title":"Enhanced Identification of Internal Casting Defects in Vehicle Wheels Using YOLO Object Detection and X-Ray Inspection","volume":"40","author":"J-D Wu","year":"2023","journal-title":"TS"},{"issue":"6","key":"pone.0357212.ref011","doi-asserted-by":"crossref","first-page":"607","DOI":"10.22214\/ijraset.2020.6098","article-title":"An Overview of You Only Look Once: Unified, Real-Time Object Detection","volume":"8","author":"HCD","year":"2020","journal-title":"IJRASET"},{"issue":"18","key":"pone.0357212.ref012","doi-asserted-by":"crossref","first-page":"7761","DOI":"10.3390\/s23187761","article-title":"An Improved YOLOv5 Algorithm for Vulnerable Road User Detection","volume":"23","author":"W Yang","year":"2023","journal-title":"Sensors (Basel)"},{"issue":"7","key":"pone.0357212.ref013","doi-asserted-by":"crossref","first-page":"7307","DOI":"10.3758\/s13428-024-02420-5","article-title":"Automatic object detection for behavioural research using YOLOv8","volume":"56","author":"F Hermens","year":"2024","journal-title":"Behav Res Methods"},{"issue":"2","key":"pone.0357212.ref014","article-title":"Flood scenarios vehicle detection algorithm based on improved YOLOv9","volume":"31","author":"J Sun","year":"2025","journal-title":"Multimedia Systems"},{"issue":"4","key":"pone.0357212.ref015","doi-asserted-by":"crossref","DOI":"10.1007\/s43452-025-01276-6","article-title":"Improved YOLOv11-EGM deep learning model for rock fragment identification","volume":"25","author":"H Meng","year":"2025","journal-title":"Arch Civ Mech Eng"},{"key":"pone.0357212.ref016","doi-asserted-by":"crossref","first-page":"102226","DOI":"10.1016\/j.displa.2022.102226","article-title":"A lightweight hand gesture recognition in complex backgrounds","volume":"74","author":"W Zhou","year":"2022","journal-title":"Displays"},{"key":"pone.0357212.ref017","doi-asserted-by":"crossref","first-page":"105066","DOI":"10.1016\/j.autcon.2023.105066","article-title":"Detection of moving objects in multi-complex environments using selective attention networks (SANet)","volume":"155","author":"J Cho","year":"2023","journal-title":"Automation in Construction"},{"issue":"17","key":"pone.0357212.ref018","doi-asserted-by":"crossref","first-page":"3527","DOI":"10.3390\/electronics13173527","article-title":"Research on Low-Light Environment Object Detection Algorithm Based on YOLO_GD","volume":"13","author":"J Li","year":"2024","journal-title":"Electronics"},{"key":"pone.0357212.ref019","doi-asserted-by":"crossref","first-page":"105484","DOI":"10.1016\/j.dsp.2025.105484","article-title":"RCPU-Net: A multi-scale multi-object segmentation model for coal and gangue under uneven lighting based on improved U-Net","volume":"168","author":"N Li","year":"2026","journal-title":"Digital Signal Processing"},{"key":"pone.0357212.ref020","doi-asserted-by":"crossref","first-page":"119494","DOI":"10.1016\/j.measurement.2025.119494","article-title":"LL-YOLO: A lightweight pavement disease detection algorithm in low-light environments","volume":"258","author":"Y Jia","year":"2026","journal-title":"Measurement"},{"key":"pone.0357212.ref021","doi-asserted-by":"crossref","first-page":"101414","DOI":"10.1016\/j.atech.2025.101414","article-title":"CD-ViT-YOLO: A lightweight Hybrid ViT-YOLO model for caged duck behaviour recognition under varying lighting conditions","volume":"12","author":"Y Gong","year":"2025","journal-title":"Smart Agricultural Technology"},{"key":"pone.0357212.ref022","doi-asserted-by":"crossref","first-page":"105355","DOI":"10.1016\/j.dsp.2025.105355","article-title":"FE-YOLO: Fourier enhancement YOLO for end-to-end object detection in low-light conditions","volume":"166","author":"Y Li","year":"2025","journal-title":"Digital Signal Processing"},{"key":"pone.0357212.ref023","doi-asserted-by":"crossref","first-page":"131778","DOI":"10.1016\/j.neucom.2025.131778","article-title":"YOLO-FOD: Lightweight object detection based on multibranch and multiscale feature fusion for adverse weather","volume":"659","author":"Y Liu","year":"2026","journal-title":"Neurocomputing"},{"key":"pone.0357212.ref024","doi-asserted-by":"crossref","first-page":"109281","DOI":"10.1016\/j.compag.2024.109281","article-title":"DHSW-YOLO: A duck flock daily behavior recognition model adaptable to bright and dark conditions","volume":"225","author":"D Xiao","year":"2024","journal-title":"Computers and Electronics in Agriculture"},{"issue":"1","key":"pone.0357212.ref025","doi-asserted-by":"crossref","first-page":"14032","DOI":"10.1038\/s41598-025-96314-x","article-title":"Research on object detection and recognition in remote sensing images based on YOLOv11","volume":"15","author":"L-H He","year":"2025","journal-title":"Sci Rep"},{"issue":"7","key":"pone.0357212.ref026","article-title":"Workpiece surface defect detection based on YOLOv11 and edge computing","volume":"20","author":"Z Wang","year":"2025","journal-title":"PLoS One"},{"issue":"22","key":"pone.0357212.ref027","first-page":"342","article-title":"Research and implementation of pork carcass cutting and localization based on improved YOLOv11n algorithm","volume":"41","author":"J Yang","year":"2025","journal-title":"Transactions of the Chinese Society of Agricultural Engineering"},{"key":"pone.0357212.ref028","doi-asserted-by":"crossref","unstructured":"Chen L, Gu L, Zheng D, Fu Y. Frequency-Adaptive Dilated Convolution for Semantic Segmentation. In: 2024 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024. 3414\u201325. https:\/\/doi.org\/10.1109\/cvpr52733.2024.00328","DOI":"10.1109\/CVPR52733.2024.00328"},{"key":"pone.0357212.ref029","article-title":"How do vision transformers work?","author":"N Park","year":"2022","journal-title":"arXiv preprint"},{"key":"pone.0357212.ref030","first-page":"78","volume-title":"MobileNetV4: Universal models for the mobile ecosystem. In: European Conference on Computer Vision","author":"D Qin","year":"2024"},{"key":"pone.0357212.ref031","doi-asserted-by":"crossref","first-page":"128686","DOI":"10.1016\/j.eswa.2025.128686","article-title":"UCIW-YOLO: Multi-category and high-precision obstacle detection model for agricultural machinery in unstructured farmland environments","volume":"294","author":"G Liu","year":"2025","journal-title":"Expert Systems with Applications"},{"key":"pone.0357212.ref032","doi-asserted-by":"crossref","unstructured":"Dai J, Qi H, Xiong Y, Li Y, Zhang G, Hu H, et al. Deformable Convolutional Networks. In: 2017 IEEE International Conference on Computer Vision (ICCV), 2017. 764\u201373. https:\/\/doi.org\/10.1109\/iccv.2017.89","DOI":"10.1109\/ICCV.2017.89"},{"key":"pone.0357212.ref033","doi-asserted-by":"crossref","unstructured":"Zhu L, Wang X, Ke Z, Zhang W, Lau R. BiFormer: Vision Transformer with Bi-Level Routing Attention. In: 2023 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2023. 10323\u201333. https:\/\/doi.org\/10.1109\/cvpr52729.2023.00995","DOI":"10.1109\/CVPR52729.2023.00995"},{"key":"pone.0357212.ref034","unstructured":"KOTO. China Traffic Police Gesture Dataset. 2019. https:\/\/www.heywhale.com\/mw\/dataset\/5de75df5ca27f8002c4cf1bb"},{"issue":"8","key":"pone.0357212.ref035","first-page":"100","article-title":"Traffic police gesture recognition based on improved YOLOX-tiny algorithm","volume":"47","author":"W Fang","year":"2024","journal-title":"Electric Measurement Technology"},{"key":"pone.0357212.ref036","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1080\/19427867.2026.2631120","article-title":"Exploring factors of E-bike rider\u2019s hit-and-run behavior of E-bike crashes on urban and rural roads using Light Gradient Boosting machine and SHapley additive exPlanations","author":"G Li","year":"2026","journal-title":"Transportation Letters"}],"container-title":["PLOS One"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0357212","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,8,28]],"date-time":"2026-08-28T17:43:24Z","timestamp":1787939004000},"score":1,"resource":{"primary":{"URL":"https:\/\/dx.plos.org\/10.1371\/journal.pone.0357212"}},"subtitle":[],"editor":[{"given":"Gen","family":"Li","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"editor"}]}],"short-title":[],"issued":{"date-parts":[[2026,8,28]]},"references-count":36,"journal-issue":{"issue":"8","published-online":{"date-parts":[[2026,8,28]]}},"URL":"https:\/\/doi.org\/10.1371\/journal.pone.0357212","relation":{},"ISSN":["1932-6203"],"issn-type":[{"value":"1932-6203","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,8,28]]}}}