{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,10]],"date-time":"2026-07-10T05:47:43Z","timestamp":1783662463313,"version":"3.55.0"},"reference-count":42,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,9,9]],"date-time":"2025-09-09T00:00:00Z","timestamp":1757376000000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:sec><jats:title>Introduction<\/jats:title><jats:p>Polymer material films are produced through extrusion machines, and their surfaces can develop micro-defects due to process and operational influences. The quantity and size of these defects significantly impact product quality.<\/jats:p><\/jats:sec><jats:sec><jats:title>Methods<\/jats:title><jats:p>As traditional machine learning defect detection methods suffer from low accuracy and poor adaptability to complex scenarios, requiring extensive effort for parameter tuning and exhibiting weak generalization capability, this paper proposes an improved YOLOv8 method to identify micro-defects on films. The approach embeds the CBAM attention mechanism into high-level networks to address feature sparsity in small target detection samples. Simultaneously, given the difficulty in obtaining large annotated datasets, we employ the Mean Teacher method for semi-supervised learning using limited labeled data. During training, the method optimizes neural network gradients through an improved loss function based on normalized Wasserstein distance (NWD), mitigating gradient instability caused by scale variations and enhancing detection accuracy for small targets. Additionally, a proposed multi-threshold mask segmentation algorithm extracts defect contours for further feature analysis.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>Experimental results demonstrate that the improved YOLOv8 algorithm achieves an 8.26% increase in mAP@0.5 compared to the baseline. It exhibits higher precision for small targets, and maintains defect detection rates exceeding 95.0% across validation data of varying image sizes, thereby meeting industrial production requirements. In generalization validation, the model demonstrates superior performance compared to traditional methods under test environments with lighting variations and environmental contamination.<\/jats:p><\/jats:sec><jats:sec><jats:title>Discussion<\/jats:title><jats:p>The improved YOLOv8 algorithm meeting the stringent requirements for high-precision small-target defect detection on polymer material film in industrial production. Future work will explore more advanced techniques to enhance model accuracy and robustness.<\/jats:p><\/jats:sec>","DOI":"10.3389\/frai.2025.1638772","type":"journal-article","created":{"date-parts":[[2025,9,9]],"date-time":"2025-09-09T05:24:56Z","timestamp":1757395496000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":4,"title":["Enhanced YOLOv8 for industrial polymer films: a semi-supervised framework for micron-scale defect detection"],"prefix":"10.3389","volume":"8","author":[{"given":"Xiaoxia","family":"Yu","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Bingyu","family":"Hu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Weifeng","family":"Jiang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jinru","family":"Wan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xinduoji","family":"Yang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Nianbo","family":"Liu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiaoyan","family":"Dong","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2025,9,9]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"040801","DOI":"10.1115\/1.4049535","article-title":"Image-based surface defect detection using deep learning: a review","volume":"21","author":"Bhatt","year":"2021","journal-title":"J. Comp. Inform. Sci. Eng"},{"key":"B2","doi-asserted-by":"crossref","first-page":"245","DOI":"10.1117\/12.2686615","article-title":"\u201cSmall object detection for mobile behavior recognition based on wasserstein distance and partial convolution,\u201d","volume-title":"Optoelectronic Imaging and Multimedia Technology X","author":"Cai","year":"2023"},{"key":"B3","doi-asserted-by":"crossref","first-page":"41","DOI":"10.1109\/ICFTIC57696.2022.10075095","article-title":"\u201cMethod for detecting surface defects of ceramic tile based on improved cascade RCNN,\u201d","volume-title":"2022 4th International Conference on Frontiers Technology of Information and Computer (ICFTIC)","author":"Cao","year":"2022"},{"key":"B4","doi-asserted-by":"crossref","first-page":"280","DOI":"10.1109\/ISCTIS51085.2021.00064","article-title":"\u201cWafer maps defect recognition based on transfer learning of handwritten pre-training network,\u201d","volume-title":"2021 International Symposium on Computer Technology and Information Science (ISCTIS)","author":"Chen","year":"2021"},{"key":"B5","doi-asserted-by":"publisher","first-page":"7657","DOI":"10.3390\/app11167657","article-title":"Surface defect detection methods for industrial products: a review","volume":"11","author":"Chen","year":"","journal-title":"Appl. Sci"},{"key":"B6","doi-asserted-by":"publisher","first-page":"108970","DOI":"10.1016\/j.compscitech.2021.108970","article-title":"Advances in mechanics of hierarchical composite materials","volume":"214","author":"Chen","year":"","journal-title":"Compos. Sci. Technol"},{"key":"B7","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1605.06409","article-title":"R-fcn: Object detection via region-based fully convolutional networks","author":"Dai","year":"2016","journal-title":"arXiv"},{"key":"B8","doi-asserted-by":"publisher","first-page":"7233","DOI":"10.1007\/s00371-025-03802-2","article-title":"MTMFNet: multi-threshold and multi-scale feature fusion network for text detection","volume":"41","author":"Dai","year":"2025","journal-title":"Visual Comp"},{"key":"B9","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1109\/PuneCon58714.2023.10450059","article-title":"\u201cEnhancing fabric integrity: Seg-YOLO-based defect detection in handloom fibers,\u201d","volume-title":"2023 IEEE Pune Section International Conference (PuneCon)","author":"Das","year":"2023"},{"key":"B10","first-page":"1440","author":"Girshick","year":"2015"},{"key":"B11","first-page":"580","article-title":"\u201cRich feature hierarchies for accurate object detection and semantic segmentation,\u201d","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition","author":"Girshick","year":"2014"},{"key":"B12","first-page":"1160","article-title":"\u201cEffective fusion factor in fpn for tiny object detection,\u201d","volume-title":"Proceedings of the IEEE\/CVF Winter Conference on Applications of Computer Vision (WACV)","author":"Gong","year":"2021"},{"key":"B13","doi-asserted-by":"publisher","first-page":"221","DOI":"10.1109\/TCSVT.2023.3284161","article-title":"Save the tiny, save the all: Hierarchical activation network for tiny object detection","volume":"34","author":"Guo","year":"2023","journal-title":"IEEE trans. Circuits Systems Video Technol"},{"key":"B14","doi-asserted-by":"publisher","first-page":"588","DOI":"10.1016\/j.apm.2017.02.015","article-title":"A new multi-threshold image segmentation approach using state transition algorithm","volume":"44","author":"Han","year":"2017","journal-title":"Appl. Math. Model"},{"key":"B15","first-page":"2961","article-title":"\u201cMask R-CNN,\u201d","volume-title":"Proceedings of the IEEE International Conference on Computer Vision","author":"He","year":"2017"},{"key":"B16","doi-asserted-by":"publisher","first-page":"316","DOI":"10.1109\/TSM.2024.3418520","article-title":"Recognition and classification of mixed defect pattern wafer map based on multi path DCNN","volume":"37","author":"Hou","year":"2024","journal-title":"IEEE Trans. Semiconduct. Manufact"},{"key":"B17","doi-asserted-by":"publisher","first-page":"113640","DOI":"10.1016\/j.compstruct.2021.113640","article-title":"Polymer composite materials: a comprehensive review","volume":"262","author":"Hsissou","year":"2021","journal-title":"Composite Struct"},{"key":"B18","doi-asserted-by":"publisher","first-page":"677","DOI":"10.3390\/machines11070677","article-title":"YOLO-v1 to YOLO-v8, the rise of YOLO and its complementary nature toward digital manufacturing and industrial defect detection","volume":"11","author":"Hussain","year":"2023","journal-title":"Machines"},{"key":"B19","doi-asserted-by":"publisher","first-page":"204","DOI":"10.1016\/j.neunet.2022.05.024","article-title":"MGLNN: semi-supervised learning via multiple graph cooperative learning neural networks","volume":"153","author":"Jiang","year":"2022","journal-title":"Neural Netw"},{"key":"B20","doi-asserted-by":"crossref","first-page":"7553","DOI":"10.1109\/CAC53003.2021.9728544","article-title":"\u201cA method of electricity meter lcd screen defect detecting based on convolutional neural network,\u201d","volume-title":"2021 China Automation Congress (CAC)","author":"Kou","year":"2021"},{"key":"B21","doi-asserted-by":"crossref","first-page":"177","DOI":"10.1109\/M2GARSS57310.2024.10537359","article-title":"\u201cDevelopment of an advanced tool for the automatic detection and correction of irregular defects in raw satellite imagery,\u201d","volume-title":"2024 IEEE Mediterranean and Middle-East Geoscience and Remote Sensing Symposium (M2GARSS)","author":"Laidi","year":"2024"},{"key":"B22","doi-asserted-by":"crossref","DOI":"10.23919\/CCC58697.2023.10240207","article-title":"\u201cGCSC-detector: a detector for photovoltaic cell defect based on deep learning,\u201d","volume-title":"2023 42nd Chinese Control Conference (CCC)","author":"Li","year":"2023"},{"key":"B23","first-page":"2980","article-title":"\u201cFocal loss for dense object detection,\u201d","volume-title":"Proceedings of the IEEE International Conference on Computer Vision","author":"Lin","year":"2017"},{"key":"B24","doi-asserted-by":"crossref","first-page":"21","DOI":"10.1007\/978-3-319-46448-0_2","article-title":"\u201cSSD: single shot multibox detector,\u201d","volume-title":"Computer Vision-ECCV 2016: 14th European Conference","author":"Liu","year":"2016"},{"key":"B25","doi-asserted-by":"publisher","first-page":"360","DOI":"10.1007\/s10015-021-00686-y","article-title":"Defect detection in wrap film product using compact convolutional neural network","volume":"26","author":"Nakashima","year":"2021","journal-title":"Artif. Life Robotics"},{"key":"B26","doi-asserted-by":"publisher","first-page":"61","DOI":"10.24425\/mper.2023.145366","article-title":"Vision-based online defect detection of polymeric film via structural quality metrics","volume":"14","author":"Rawashedeh","year":"2023","journal-title":"Managem. Prod. Eng. Rev"},{"key":"B27","first-page":"779","article-title":"\u201cYou only look once: unified, real-time object detection,\u201d","volume-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition","author":"Redmon","year":"2016"},{"key":"B28","doi-asserted-by":"publisher","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":"Ren","year":"2016","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell"},{"key":"B29","doi-asserted-by":"publisher","first-page":"1995","DOI":"10.1007\/s10845-021-01878-w","article-title":"Automated surface defect detection framework using machine vision and convolutional neural networks","volume":"34","author":"Singh","year":"2023","journal-title":"J. Intell. Manuf"},{"key":"B30","doi-asserted-by":"crossref","first-page":"220","DOI":"10.1109\/ICAIDT62617.2024.00055","article-title":"\u201cAn AHP-based defect detection algorithm study for e-paper pockmarks detection,\u201d","volume-title":"2024 International Conference on Artificial Intelligence and Digital Technology (ICAIDT)","author":"Sun","year":"2024"},{"key":"B31","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2110.13389","article-title":"A normalized gaussian wasserstein distance for tiny object detection","author":"Wang","year":"","journal-title":"arXiv"},{"key":"B32","first-page":"3791","article-title":"\u201cTiny object detection in aerial images,\u201d","volume-title":"2020 25th International Conference on Pattern Recognition (ICPR)","author":"Wang","year":""},{"key":"B33","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2205.07246","article-title":"Freematch: Self-adaptive thresholding for semi-supervised learning","author":"Wang","year":"2022","journal-title":"arXiv"},{"key":"B34","doi-asserted-by":"publisher","first-page":"118665","DOI":"10.1016\/j.eswa.2022.118665","article-title":"Tiny object detection with context enhancement and feature purification","volume":"211","author":"Xiao","year":"2023","journal-title":"Expert Syst. Appl"},{"key":"B35","doi-asserted-by":"publisher","first-page":"114970","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"},{"key":"B36","doi-asserted-by":"publisher","first-page":"728","DOI":"10.1007\/s11760-025-04335-w","article-title":"CSLNet: an enhanced YOLOv8-based approach to defect surface foreign objects in lyophilized powder","volume":"19","author":"Xu","year":"","journal-title":"Signal, Image Video Proc"},{"key":"B37","doi-asserted-by":"publisher","first-page":"625","DOI":"10.1007\/s11227-025-07135-8","article-title":"A vision-based inspection system for pharmaceutical production line","volume":"81","author":"Xu","year":"","journal-title":"J. Supercomput"},{"key":"B38","first-page":"11525","article-title":"\u201cDash: semi-supervised learning with dynamic thresholding,\u201d","volume-title":"International Conference on Machine Learning","author":"Xu","year":"2021"},{"key":"B39","first-page":"1840","article-title":"Flexmatch: Boosting semi-supervised learning with curriculum pseudo labeling","volume":"34","author":"Zhang","year":"2021","journal-title":"Adv. Neural Inf. Process. Syst"},{"key":"B40","doi-asserted-by":"publisher","first-page":"106390","DOI":"10.1016\/j.engappai.2023.106390","article-title":"Idd-net: Industrial defect detection method based on deep-learning","volume":"123","author":"Zhang","year":"2023","journal-title":"Eng. Appl. Artif. Intell"},{"key":"B41","first-page":"16965","article-title":"\u201cDetrs beat yolos on real-time object detection,\u201d","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Zhao","year":"2024"},{"key":"B42","first-page":"14471","article-title":"\u201cSimMatch: semi-supervised learning with similarity matching,\u201d","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Zheng","year":"2022"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1638772\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,9,9]],"date-time":"2025-09-09T05:25:01Z","timestamp":1757395501000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1638772\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,9,9]]},"references-count":42,"alternative-id":["10.3389\/frai.2025.1638772"],"URL":"https:\/\/doi.org\/10.3389\/frai.2025.1638772","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,9,9]]},"article-number":"1638772"}}