{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,21]],"date-time":"2026-08-21T01:16:41Z","timestamp":1787275001225,"version":"build-2736575974"},"reference-count":41,"publisher":"Wiley","issue":"1","license":[{"start":{"date-parts":[[2025,11,30]],"date-time":"2025-11-30T00:00:00Z","timestamp":1764460800000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/onlinelibrary.wiley.com\/termsAndConditions#vor"},{"start":{"date-parts":[[2025,11,30]],"date-time":"2025-11-30T00:00:00Z","timestamp":1764460800000},"content-version":"tdm","delay-in-days":0,"URL":"http:\/\/doi.wiley.com\/10.1002\/tdm_license_1.1"}],"content-domain":{"domain":["nyaspubs.onlinelibrary.wiley.com"],"crossmark-restriction":true},"short-container-title":["Ann NY Acad Sci"],"published-print":{"date-parts":[[2026,2]]},"abstract":"<jats:title>ABSTRACT<\/jats:title>\n                  <jats:p>Bearing surface defect detection is critical for industrial equipment reliability, but existing deep learning methods suffer from low accuracy for small targets, high computational complexity, and limited edge device deployment. This paper proposes an efficient defect detection algorithm based on the StarNet\u2010MEIS\u2010FDConv\u2010detection transformer (SMF\u2010DETR). The algorithm employs element\u2010level multiplication operations in the backbone network to achieve high\u2010dimensional feature mapping, effectively reducing computational complexity while improving feature extraction capability. The multiscale edge information selection mechanism processes features at different resolutions simultaneously to improve small defect detection. Frequency domain dynamic convolution adapts to different frequency components for optimal feature extraction while maintaining computational efficiency. Experiments on custom bearing defect datasets show that SMF\u2010DETR achieves 96.2% mean average precision@50 (mAP@50) and 98.1% accuracy, improving baseline performance by 3.1% and 2.9%, respectively. The model also reduces computational cost by 57.7% and model size by 37.1%. Processing speeds reach 97.3 frames per second (FPS) on desktop systems and 58.1 FPS on embedded RK3588 platforms, meeting industrial real\u2010time detection requirements. Finally, experimental validation was conducted on the publicly available bearing defect\u2010detection dataset and the PASCAL visual object classes dataset, demonstrating the algorithm's versatility and generalization capabilities.<\/jats:p>","DOI":"10.1111\/nyas.70156","type":"journal-article","created":{"date-parts":[[2025,11,30]],"date-time":"2025-11-30T13:39:42Z","timestamp":1764509982000},"update-policy":"https:\/\/doi.org\/10.1002\/crossmark_policy","source":"Crossref","is-referenced-by-count":4,"title":["SMF\u2010DETR: An Efficient Lightweight Detection Transformer for Real\u2010Time Bearing Surface Defect Detection"],"prefix":"10.1111","volume":"1556","author":[{"ORCID":"https:\/\/orcid.org\/0009-0007-3549-9402","authenticated-orcid":false,"given":"Min","family":"Gao","sequence":"first","affiliation":[{"name":"Department of Mechanical and Electrical Engineering Shanxi Institute of Energy  Jinzhong China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiaoping","family":"Kang","sequence":"additional","affiliation":[{"name":"Department of Mechanical and Electrical Engineering Shanxi Institute of Energy  Jinzhong China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kun","family":"Zhou","sequence":"additional","affiliation":[{"name":"Department of Energy Chemistry and Materials Engineering Shanxi Institute of Energy  Jinzhong China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Teng","family":"Xie","sequence":"additional","affiliation":[{"name":"Department of Planning and Finance Shanxi Medical University  Jinzhong China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"311","published-online":{"date-parts":[[2025,11,30]]},"reference":[{"key":"e_1_2_11_2_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.isatra.2023.08.014"},{"key":"e_1_2_11_3_1","doi-asserted-by":"publisher","DOI":"10.1016\/S0262-8856(02)00152-X"},{"key":"e_1_2_11_4_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.optlaseng.2018.01.010"},{"key":"e_1_2_11_5_1","doi-asserted-by":"crossref","unstructured":"R.Girshick J.Donahue T.Darrell andJ.Malik \u201cRich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation \u201d in2014 IEEE Conference on Computer Vision and Pattern Recognition Columbus OH USA 23\u201328 June (2014) 580\u2013587.","DOI":"10.1109\/CVPR.2014.81"},{"key":"e_1_2_11_6_1","doi-asserted-by":"publisher","DOI":"10.1007\/s10489-020-01665-9"},{"key":"e_1_2_11_7_1","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2016.2577031"},{"key":"e_1_2_11_8_1","doi-asserted-by":"publisher","DOI":"10.1049\/iet-cvi.2019.0711"},{"key":"e_1_2_11_9_1","doi-asserted-by":"publisher","DOI":"10.3390\/app132413044"},{"key":"e_1_2_11_10_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.ijepes.2022.108277"},{"key":"e_1_2_11_11_1","unstructured":"L.Wenyu X.Shangliang Z.Yian et\u00a0al. DETRs Beat YOLOs on Real\u2010Time Object Detection. (2023) https:\/\/doi.org\/10.48550\/arXiv.2304.08069."},{"key":"e_1_2_11_12_1","unstructured":"M.Xu D.Xiyang B.Yue W.Yizhou andF.Yun Rewrite the Stars. (2024) https:\/\/doi.org\/10.48550\/arXiv.2403.19967."},{"key":"e_1_2_11_13_1","unstructured":"C.Linwei G.Lin L.Liang Y.Chenggang andF.Ying Frequency Dynamic Convolution for Dense Image Prediction. (2025) https:\/\/doi.org\/10.48550\/arXiv.2503.18783."},{"key":"e_1_2_11_14_1","doi-asserted-by":"crossref","unstructured":"W.Li Y.Zheng andX.Wang \u201cStudy on End Face Defect Detection Method of Bearing Inner Ring \u201d in2018 5th Asia\u2010Pacific World Congress on Computer Science and Engineering (APWC on CSE) Nadi Fiji 10\u201312 December (2018) 172\u2013177.","DOI":"10.1109\/APWConCSE.2018.00035"},{"key":"e_1_2_11_15_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.measurement.2011.12.018"},{"key":"e_1_2_11_16_1","doi-asserted-by":"crossref","unstructured":"X.Li J.Wang F.Xu andJ.Song \u201cImprovement of YOLOv3 Algorithm in Workpiece Detection \u201d in2019 IEEE 9th Annual International Conference on CYBER Technology in Automation Control and Intelligent Systems (CYBER) Suzhou China 29 July\u20132 August (IEEE 2019) 1063\u20131068.","DOI":"10.1109\/CYBER46603.2019.9066490"},{"key":"e_1_2_11_17_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.measurement.2023.112776"},{"key":"e_1_2_11_18_1","doi-asserted-by":"publisher","DOI":"10.1109\/JSTARS.2022.3181303"},{"key":"e_1_2_11_19_1","doi-asserted-by":"crossref","unstructured":"J. A.Zhao andM. P.Jia \u201cFeature Extraction With Global\u2010Locally Preserving Projections Based on Curvelet Transform \u201d in Proc. Electronics Electrical Engineering and Information Science (World Scientific 2016) 345\u2013353.","DOI":"10.1142\/9789814740135_0036"},{"key":"e_1_2_11_20_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.eswa.2023.122365"},{"key":"e_1_2_11_21_1","doi-asserted-by":"publisher","DOI":"10.3390\/app11041825"},{"key":"e_1_2_11_22_1","doi-asserted-by":"publisher","DOI":"10.3390\/s23177437"},{"key":"e_1_2_11_23_1","doi-asserted-by":"publisher","DOI":"10.3390\/s22093467"},{"key":"e_1_2_11_24_1","doi-asserted-by":"publisher","DOI":"10.3390\/electronics13132614"},{"key":"e_1_2_11_25_1","doi-asserted-by":"crossref","unstructured":"N.Carion F.Massa G.Synnaeve N.Usunier A.Kirillov andS.Zagoruyko \u201cEnd\u2010to\u2010End Object Detection With Transformers \u201d inComputer Vision\u2014ECCV 2020: 16th European Conference Glasgow UK August 23\u201328 2020 Proceedings Part I (Springer\u2010Verlag 2020) 213\u2013229.","DOI":"10.1007\/978-3-030-58452-8_13"},{"key":"e_1_2_11_26_1","doi-asserted-by":"crossref","unstructured":"Q.Chen X.Chen J.Wang et\u00a0al. \u201cGroup DETR: Fast DETR Training With Group\u2010Wise One\u2010to\u2010Many Assignment \u201d in2023 IEEE\/CVF International Conference on Computer Vision (ICCV) Paris France 1\u20136 October 2023 (IEEE 2023) 6610\u20136619.","DOI":"10.1109\/ICCV51070.2023.00610"},{"key":"e_1_2_11_27_1","doi-asserted-by":"crossref","unstructured":"Z.Zong G.Song andY.Liu \u201cDETRs With Collaborative Hybrid Assignments Training \u201d in2023 IEEE\/CVF International Conference on Computer Vision (ICCV) Paris France 1\u20136 October 2023 (IEEE 2023) 6725\u20136735.","DOI":"10.1109\/ICCV51070.2023.00621"},{"key":"e_1_2_11_28_1","unstructured":"S.Liu F.Li H.Zhang et\u00a0al. DAB\u2010DETR: Dynamic Anchor Boxes Are Better Queries for DETR. (2022) https:\/\/doi.org\/10.48550\/arXiv.2201.12329."},{"key":"e_1_2_11_29_1","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2023.3335410"},{"key":"e_1_2_11_30_1","unstructured":"H.Zhang F.Li S.Liu et\u00a0al. DINO: DETR With Improved Denoising Anchor Boxes for End\u2010to\u2010End Object Detection. (2022) https:\/\/doi.org\/10.48550\/arXiv.2203.03605."},{"key":"e_1_2_11_31_1","unstructured":"Z.Yao J.Ai B.Li andC.Zhang Efficient DETR: Improving End\u2010to\u2010End Object Detector With Dense Prior. (2021) https:\/\/doi.org\/10.48550\/arXiv.2104.01318."},{"key":"e_1_2_11_32_1","unstructured":"Y.Peng H.Li P.Wu Y.Zhang X.Sun andF.Wu D\u2010FINE: Redefine Regression Task in DETRs as Fine\u2010Grained Distribution Refinement. (2024) https:\/\/doi.org\/10.48550\/arXiv.2410.13842."},{"key":"e_1_2_11_33_1","doi-asserted-by":"crossref","unstructured":"S.Huang Z.Lu X.Cun Y.Yu X.Zhou andX.Shen \u201cDEIM: DETR With Improved Matching for Fast Convergence \u201d in2025 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Nashville TN USA 10\u201317 June 2025 (IEEE 2025) 15162\u201315171.","DOI":"10.1109\/CVPR52734.2025.01412"},{"key":"e_1_2_11_34_1","doi-asserted-by":"crossref","unstructured":"S.Liu E. J.Davison andJ.Andrew End\u2010to\u2010End Multi\u2010Task Learning With Attention. (2018) https:\/\/doi.org\/10.48550\/arXiv.1803.10704.","DOI":"10.1109\/CVPR.2019.00197"},{"key":"e_1_2_11_35_1","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2021.3070203"},{"key":"e_1_2_11_36_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.eswa.2024.124551"},{"key":"e_1_2_11_37_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.infrared.2024.105223"},{"key":"e_1_2_11_38_1","doi-asserted-by":"publisher","DOI":"10.1088\/2631-8695\/adcdc3"},{"key":"e_1_2_11_39_1","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1109\/TIM.2024.3472791","article-title":"BED\u2010YOLO: An Enhanced YOLOv8 for High\u2010Precision Real\u2010Time Bearing Defect Detection","volume":"73","author":"Han T.","year":"2024","journal-title":"IEEE Transactions on Instrumentation and Measurement"},{"key":"e_1_2_11_40_1","doi-asserted-by":"publisher","DOI":"10.1088\/1402-4896\/ad9ef1"},{"key":"e_1_2_11_41_1","doi-asserted-by":"publisher","DOI":"10.1371\/journal.pone.0310007"},{"key":"e_1_2_11_42_1","first-page":"1","article-title":"Bearing Defect Detection With Unsupervised Neural Networks","volume":"10","author":"Xu J.","year":"2021","journal-title":"Shock and Vibration"}],"container-title":["Annals of the New York Academy of Sciences"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/nyaspubs.onlinelibrary.wiley.com\/doi\/pdf\/10.1111\/nyas.70156","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/nyaspubs.onlinelibrary.wiley.com\/doi\/full-xml\/10.1111\/nyas.70156","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/nyaspubs.onlinelibrary.wiley.com\/doi\/pdf\/10.1111\/nyas.70156","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,29]],"date-time":"2026-04-29T16:15:24Z","timestamp":1777479324000},"score":1,"resource":{"primary":{"URL":"https:\/\/nyaspubs.onlinelibrary.wiley.com\/doi\/10.1111\/nyas.70156"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,11,30]]},"references-count":41,"journal-issue":{"issue":"1","published-print":{"date-parts":[[2026,2]]}},"alternative-id":["10.1111\/nyas.70156"],"URL":"https:\/\/doi.org\/10.1111\/nyas.70156","archive":["Portico"],"relation":{"has-review":[{"id-type":"doi","id":"10.1111\/NYAS.70156\/v1\/review3","asserted-by":"object"},{"id-type":"doi","id":"10.1111\/NYAS.70156\/v1\/review2","asserted-by":"object"},{"id-type":"doi","id":"10.1111\/NYAS.70156\/v1\/review1","asserted-by":"object"},{"id-type":"doi","id":"10.1111\/NYAS.70156\/v1\/decision1","asserted-by":"object"}]},"ISSN":["0077-8923","1749-6632"],"issn-type":[{"value":"0077-8923","type":"print"},{"value":"1749-6632","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,11,30]]},"assertion":[{"value":"2025-11-30","order":3,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}],"article-number":"e70156"}}