{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,4]],"date-time":"2026-05-04T05:57:30Z","timestamp":1777874250399,"version":"3.51.4"},"reference-count":47,"publisher":"Elsevier BV","license":[{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/tdm\/userlicense\/1.0\/"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/www.elsevier.com\/legal\/tdmrep-license"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-017"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-012"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2026,7,1]],"date-time":"2026-07-01T00:00:00Z","timestamp":1782864000000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-004"}],"funder":[{"DOI":"10.13039\/501100004826","name":"Beijing Natural Science Foundation","doi-asserted-by":"publisher","award":["L241078"],"award-info":[{"award-number":["L241078"]}],"id":[{"id":"10.13039\/501100004826","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100005022","name":"Beijing Jiaotong University","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100005022","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["elsevier.com","sciencedirect.com"],"crossmark-restriction":true},"short-container-title":["Neurocomputing"],"published-print":{"date-parts":[[2026,7]]},"DOI":"10.1016\/j.neucom.2026.133643","type":"journal-article","created":{"date-parts":[[2026,4,11]],"date-time":"2026-04-11T08:28:11Z","timestamp":1775896091000},"page":"133643","update-policy":"https:\/\/doi.org\/10.1016\/elsevier_cm_policy","source":"Crossref","is-referenced-by-count":0,"special_numbering":"C","title":["MMLight: Lightweight unsupervised multi-modal feature fusion method for track defect detection"],"prefix":"10.1016","volume":"685","author":[{"ORCID":"https:\/\/orcid.org\/0009-0000-2456-3850","authenticated-orcid":false,"given":"Yang","family":"Gao","sequence":"first","affiliation":[]},{"given":"Yong","family":"Qin","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1611-9703","authenticated-orcid":false,"given":"Zhiwei","family":"Cao","sequence":"additional","affiliation":[]},{"given":"Lirong","family":"Lian","sequence":"additional","affiliation":[]},{"given":"Yaguan","family":"Wang","sequence":"additional","affiliation":[]},{"given":"Linlin","family":"Kou","sequence":"additional","affiliation":[]},{"ORCID":"https:\/\/orcid.org\/0009-0005-9655-4239","authenticated-orcid":false,"given":"Genwang","family":"Peng","sequence":"additional","affiliation":[]}],"member":"78","reference":[{"issue":"4","key":"10.1016\/j.neucom.2026.133643_bib0005","doi-asserted-by":"crossref","first-page":"890","DOI":"10.1049\/hve2.70044","article-title":"GSLI-RTMdet: an automatic nondestructive detection method for internal defects in gas-insulated switchgear X-DR images","volume":"10","author":"Liu","year":"2025","journal-title":"High Voltage"},{"key":"10.1016\/j.neucom.2026.133643_bib0010","doi-asserted-by":"crossref","DOI":"10.1016\/j.measurement.2024.114970","article-title":"ESMNet: an enhanced YOLOv7-based approach to detect surface defects in precision metal workpieces","volume":"235","author":"Xu","year":"2024","journal-title":"Measurement"},{"issue":"4","key":"10.1016\/j.neucom.2026.133643_bib0015","doi-asserted-by":"crossref","first-page":"1691","DOI":"10.1080\/10589759.2024.2395363","article-title":"UAV based defect detection and fault diagnosis for static and rotating wind turbine blade: a review","volume":"40","author":"Zhang","year":"2025","journal-title":"Nondestruct. Test. Eval."},{"issue":"7","key":"10.1016\/j.neucom.2026.133643_bib0020","first-page":"266","article-title":"Research on active safety methodologies for intelligent railway systems","volume":"26","author":"Qin","year":"2022","journal-title":"Engineering"},{"issue":"7","key":"10.1016\/j.neucom.2026.133643_bib0025","doi-asserted-by":"crossref","first-page":"6427","DOI":"10.1109\/TITS.2024.3412170","article-title":"Railway intrusion detection based on machine vision: a survey, challenges, and perspectives","volume":"25","author":"Cao","year":"2024","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"issue":"13","key":"10.1016\/j.neucom.2026.133643_bib0030","doi-asserted-by":"crossref","first-page":"21157","DOI":"10.1109\/JSEN.2024.3402730","article-title":"Research on Real-Time Detection System of Rail Surface Defects Based on Deep Learning","volume":"24","author":"Wang","year":"2024","journal-title":"IEEE Sens. J."},{"key":"10.1016\/j.neucom.2026.133643_bib0035","doi-asserted-by":"crossref","DOI":"10.1016\/j.autcon.2021.103596","article-title":"Real-time railroad track components inspection based on the improved YOLOv4 framework","volume":"125","author":"Guo","year":"2021","journal-title":"Autom. Constr."},{"key":"10.1016\/j.neucom.2026.133643_bib0040","doi-asserted-by":"crossref","first-page":"196249","DOI":"10.1109\/ACCESS.2024.3466559","article-title":"YOLOv8n-RSDD: a high-performance low-complexity rail surface defect detection network","volume":"12","author":"Fang","year":"2024","journal-title":"IEEE Access"},{"key":"10.1016\/j.neucom.2026.133643_bib0045","doi-asserted-by":"crossref","DOI":"10.1016\/j.engappai.2023.105965","article-title":"MSRConvNet: classification of railway track defects using multi-scale residual convolutional neural network","volume":"121","author":"Acikgoz","year":"2023","journal-title":"Eng. Appl. Artif. Intell."},{"key":"10.1016\/j.neucom.2026.133643_bib0050","series-title":"Proceedings of Asian Conference on Computer Vision","first-page":"622","article-title":"Ganomaly: semi-supervised anomaly detection via adversarial training","author":"Akcay","year":"2018"},{"key":"10.1016\/j.neucom.2026.133643_bib0055","series-title":"IEEE\/CVF International Conference on Computer Vision","first-page":"8310","article-title":"A discriminatively trained reconstruction embedding for surface anomaly detection","author":"Zavrtanik","year":"2021"},{"key":"10.1016\/j.neucom.2026.133643_bib0060","series-title":"IEEE Winter Conference on Applications of Computer Vision","first-page":"1819","article-title":"CFLOW-AD: real-time unsupervised anomaly detection with localization via conditional normalizing flows","author":"Gudovskiy","year":"2022"},{"key":"10.1016\/j.neucom.2026.133643_bib0065","series-title":"2022 IEEE\/CVF Conference on Computer Vision and Pattern Recognition","first-page":"9727","article-title":"Anomaly detection via reverse distillation from one-class embedding","author":"Deng","year":"2022"},{"key":"10.1016\/j.neucom.2026.133643_bib0070","doi-asserted-by":"crossref","first-page":"78446","DOI":"10.1109\/ACCESS.2022.3193699","article-title":"CFA: coupled-hypersphere-based feature adaptation for target-oriented anomaly localization","volume":"10","author":"Lee","year":"2022","journal-title":"IEEE Access"},{"key":"10.1016\/j.neucom.2026.133643_bib0075","series-title":"2022 IEEE\/CVF Conference on Computer Vision and Pattern Recognition","first-page":"14318","article-title":"Towards total recall in industrial anomaly detection","author":"Roth","year":"2022"},{"key":"10.1016\/j.neucom.2026.133643_bib0080","series-title":"IEEE Winter Conference on Applications of Computer Vision","first-page":"127","article-title":"EfficientAD: accurate visual anomaly detection at millisecond-level latencies","author":"Batzner","year":"2024"},{"key":"10.1016\/j.neucom.2026.133643_bib0085","doi-asserted-by":"crossref","DOI":"10.1016\/j.neucom.2025.132383","article-title":"A review of deep learning based anomaly detection","volume":"668","author":"Shi","year":"2026","journal-title":"Neurocomputing"},{"key":"10.1016\/j.neucom.2026.133643_bib0090","first-page":"1","article-title":"Segmented-aggregated framework with internal-external constraint mechanism for unsupervised track anomaly detection","volume":"74","author":"Gao","year":"2025","journal-title":"IEEE Trans. Instrum. Meas."},{"key":"10.1016\/j.neucom.2026.133643_bib0095","series-title":"HFD-Net: a benchmark framework of foreign object detection for high-speed train","volume":"173","author":"Lin","year":"2026"},{"key":"10.1016\/j.neucom.2026.133643_bib0100","doi-asserted-by":"crossref","DOI":"10.1016\/j.patcog.2025.111699","article-title":"Unsupervised anomaly detection with a temporal continuation, confidence-aware VAE-GAN","volume":"166","author":"Xing","year":"2025","journal-title":"Pattern Recognit."},{"issue":"2","key":"10.1016\/j.neucom.2026.133643_bib0105","doi-asserted-by":"crossref","first-page":"1812","DOI":"10.1109\/JSEN.2023.3336962","article-title":"Railway fastener anomaly detection via multi-sensor fusion and self driven loss reweighting","volume":"24","author":"Gao","year":"2024","journal-title":"IEEE Sens. J."},{"issue":"4","key":"10.1016\/j.neucom.2026.133643_bib0110","doi-asserted-by":"crossref","first-page":"2695","DOI":"10.1109\/TSMC.2025.3526234","article-title":"Hybrid knowledge distillation network for RGB-D Co-salient object detection","volume":"55","author":"Tu","year":"2025","journal-title":"IEEE Trans. Syst. Man Cybern. Syst."},{"key":"10.1016\/j.neucom.2026.133643_bib0115","author":"Horwitz"},{"key":"10.1016\/j.neucom.2026.133643_bib0120","series-title":"IEEE Winter Conference on Applications of Computer Vision","first-page":"2591","article-title":"Asymmetric student-teacher networks for industrial anomaly detection","author":"Rudolph","year":"2023"},{"key":"10.1016\/j.neucom.2026.133643_bib0125","series-title":"International Conference on Image, Signal Processing, and Pattern Recognition (ISPP 2024)","first-page":"767","article-title":"Dual-Branch reconstruction network for industrial anomaly detection with RGB-D Data","author":"Bi","year":"2024"},{"key":"10.1016\/j.neucom.2026.133643_bib0130","author":"Chen"},{"key":"10.1016\/j.neucom.2026.133643_bib0135","doi-asserted-by":"crossref","DOI":"10.1016\/j.neucom.2025.129408","article-title":"LPFSTNet: a lightweight and parameter-free head attention-based student-teacher network for fast 3D industrial anomaly detection","volume":"623","author":"Cheng","year":"2025","journal-title":"Neurocomputing"},{"issue":"1","key":"10.1016\/j.neucom.2026.133643_bib0140","doi-asserted-by":"crossref","first-page":"133","DOI":"10.1109\/TITS.2023.3305300","article-title":"Fast detection of railway fastener using a new lightweight network Op-YOLOv4-Tiny","volume":"25","author":"Liu","year":"2024","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"key":"10.1016\/j.neucom.2026.133643_bib0145","first-page":"1","article-title":"Real-time defect detection of track components: considering class imbalance and subtle difference between classes","volume":"70","author":"Tu","year":"2021","journal-title":"IEEE Trans. Instrum. Meas."},{"issue":"13","key":"10.1016\/j.neucom.2026.133643_bib0150","doi-asserted-by":"crossref","first-page":"15010","DOI":"10.1109\/JSEN.2023.3280117","article-title":"Automatic rail surface defect inspection using the pixelwise semantic segmentation model","volume":"23","author":"Guo","year":"2023","journal-title":"IEEE Sens. J."},{"key":"10.1016\/j.neucom.2026.133643_bib0155","series-title":"2025 IEEE 5th International Conference on Power, Electronics and Computer Applications (ICPECA)","first-page":"73","article-title":"Detection of high-speed railway track bed foreign object based on contrastive learning and adversarial training","author":"Fan","year":"2025"},{"key":"10.1016\/j.neucom.2026.133643_bib0160","doi-asserted-by":"crossref","first-page":"86646","DOI":"10.1109\/ACCESS.2021.3087705","article-title":"Modeling and optimization of semantic segmentation for track bed foreign object based on attention mechanism","volume":"9","author":"Song","year":"2021","journal-title":"IEEE Access"},{"key":"10.1016\/j.neucom.2026.133643_bib0165","first-page":"1","article-title":"Automatic railroad track components inspection using hybrid deep learning framework","volume":"72","author":"Wu","year":"2023","journal-title":"IEEE Trans. Instrum. Meas."},{"issue":"8","key":"10.1016\/j.neucom.2026.133643_bib0170","doi-asserted-by":"crossref","DOI":"10.3390\/jimaging10080192","article-title":"RailTrack-DaViT: a vision transformer-based approach for automated railway track defect detection","volume":"10","author":"Phaphuangwittayakul","year":"2024","journal-title":"J. Imaging"},{"issue":"21","key":"10.1016\/j.neucom.2026.133643_bib0175","doi-asserted-by":"crossref","first-page":"3349","DOI":"10.3390\/math12213349","article-title":"Improved real-time detection transformer-based rail fastener defect detection algorithm","volume":"12","author":"Song","year":"2024","journal-title":"Mathematics"},{"issue":"6","key":"10.1016\/j.neucom.2026.133643_bib0180","doi-asserted-by":"crossref","first-page":"8894","DOI":"10.1109\/TITS.2025.3540846","article-title":"Multi-scale CNN-transformer hybrid network for rail fastener defect detection","volume":"26","author":"He","year":"2025","journal-title":"IEEE Trans. Intell. Transp. Syst."},{"issue":"4","key":"10.1016\/j.neucom.2026.133643_bib0185","doi-asserted-by":"crossref","first-page":"5025","DOI":"10.1109\/JSEN.2023.3348118","article-title":"Unified model based on reinforced feature reconstruction for metro track anomaly detection","volume":"24","author":"Duan","year":"2024","journal-title":"IEEE Sens. J."},{"issue":"13","key":"10.1016\/j.neucom.2026.133643_bib0190","doi-asserted-by":"crossref","first-page":"3408","DOI":"10.3390\/rs15133408","article-title":"Multi-branch deep learning framework for land scene classification in satellite imagery","volume":"15","author":"Khan","year":"2023","journal-title":"Remote Sens."},{"key":"10.1016\/j.neucom.2026.133643_bib0195","doi-asserted-by":"crossref","first-page":"1907","DOI":"10.1007\/s13369-022-07082-z","article-title":"DSMSA-Net: deep spatial and multi-scale attention network for road extraction in high spatial resolution satellite images","volume":"48","author":"Khan","year":"2023","journal-title":"Arab. J. Sci. Eng."},{"issue":"6","key":"10.1016\/j.neucom.2026.133643_bib0200","doi-asserted-by":"crossref","first-page":"4874","DOI":"10.1109\/TMECH.2022.3167412","article-title":"Collaborative learning attention network based on RGB image and depth image for surface defect inspection of no-service rail","volume":"27","author":"Wang","year":"2022","journal-title":"IEEE\/ASME Trans. Mechatron."},{"key":"10.1016\/j.neucom.2026.133643_bib0205","first-page":"1","article-title":"FHENet: lightweight feature hierarchical exploration network for real-time rail surface defect inspection in RGB-D images","volume":"72","author":"Zhou","year":"2023","journal-title":"IEEE Trans. Instrum. Meas."},{"key":"10.1016\/j.neucom.2026.133643_bib0210","series-title":"Proceedings of the 17th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, Online","first-page":"202","article-title":"The MVTec 3D-ad dataset for unsupervised 3D anomaly detection and localization","author":"Bergmann","year":"2022"},{"key":"10.1016\/j.neucom.2026.133643_bib0215","series-title":"2022 IEEE International Conference on Image Processing (ICIP), Bordeaux, France","first-page":"1706","article-title":"Anomalib: a deep learning library for anomaly detection","author":"Akcay","year":"2022"},{"key":"10.1016\/j.neucom.2026.133643_bib0220","series-title":"2024 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","first-page":"17234","article-title":"Multimodal industrial anomaly detection by crossmodal feature mapping","author":"Costanzino","year":"2024"},{"key":"10.1016\/j.neucom.2026.133643_bib0225","series-title":"2025 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)","first-page":"9984","article-title":"Multi-Sensor object anomaly detection: unifying appearance, geometry, and internal properties","author":"Li","year":"2025"},{"key":"10.1016\/j.neucom.2026.133643_bib0230","series-title":"Proceedings of the 33rd ACM International Conference on Multimedia","first-page":"1579","article-title":"BridgeNet: a unified multimodal framework for bridging 2D and 3D industrial anomaly detection","author":"Xiang","year":"2025"},{"key":"10.1016\/j.neucom.2026.133643_bib0235","series-title":"2023 IEEE\/CVF Conference on Computer Vision and Pattern Recognition","first-page":"8032","article-title":"Multimodal industrial anomaly detection via hybrid fusion","author":"Wang","year":"2023"}],"container-title":["Neurocomputing"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S0925231226010404?httpAccept=text\/xml","content-type":"text\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/api.elsevier.com\/content\/article\/PII:S0925231226010404?httpAccept=text\/plain","content-type":"text\/plain","content-version":"vor","intended-application":"text-mining"}],"deposited":{"date-parts":[[2026,4,30]],"date-time":"2026-04-30T19:25:59Z","timestamp":1777577159000},"score":1,"resource":{"primary":{"URL":"https:\/\/linkinghub.elsevier.com\/retrieve\/pii\/S0925231226010404"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7]]},"references-count":47,"alternative-id":["S0925231226010404"],"URL":"https:\/\/doi.org\/10.1016\/j.neucom.2026.133643","relation":{},"ISSN":["0925-2312"],"issn-type":[{"value":"0925-2312","type":"print"}],"subject":[],"published":{"date-parts":[[2026,7]]},"assertion":[{"value":"Elsevier","name":"publisher","label":"This article is maintained by"},{"value":"MMLight: Lightweight unsupervised multi-modal feature fusion method for track defect detection","name":"articletitle","label":"Article Title"},{"value":"Neurocomputing","name":"journaltitle","label":"Journal Title"},{"value":"https:\/\/doi.org\/10.1016\/j.neucom.2026.133643","name":"articlelink","label":"CrossRef DOI link to publisher maintained version"},{"value":"article","name":"content_type","label":"Content Type"},{"value":"\u00a9 2026 Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.","name":"copyright","label":"Copyright"}],"article-number":"133643"}}