{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,16]],"date-time":"2026-06-16T14:18:42Z","timestamp":1781619522334,"version":"3.54.5"},"reference-count":39,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2023,2,14]],"date-time":"2023-02-14T00:00:00Z","timestamp":1676332800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["51875380"],"award-info":[{"award-number":["51875380"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["51975394"],"award-info":[{"award-number":["51975394"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["52105526"],"award-info":[{"award-number":["52105526"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["61903269"],"award-info":[{"award-number":["61903269"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100002858","name":"China Postdoctoral Science Foundation","doi-asserted-by":"publisher","award":["2020M671383"],"award-info":[{"award-number":["2020M671383"]}],"id":[{"id":"10.13039\/501100002858","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100002858","name":"China Postdoctoral Science Foundation","doi-asserted-by":"publisher","award":["2020M681517"],"award-info":[{"award-number":["2020M681517"]}],"id":[{"id":"10.13039\/501100002858","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Neurorobot."],"abstract":"<jats:p>Surface defect detection is an important technique to realize product quality inspection. In this study, we develop an innovative multi-scale pooling convolutional neural network to accomplish high-accuracy steel surface defect classification. The model was built based on SqueezeNet, and experiments were carried out on the NEU noise-free and noisy testing set. Class activation map visualization proves that the multi-scale pooling model can accurately capture the defect location at multiple scales, and the defect feature information at different scales can complement and reinforce each other to obtain more robust results. Through T-SNE visualization analysis, it is found that the classification results of this model have large inter-class distance and small intra-class distance, indicating that this model has high reliability and strong generalization ability. In addition, the model is small in size (3MB) and runs at up to 130FPS on an NVIDIA 1080Ti GPU, making it suitable for applications with high real-time requirements.<\/jats:p>","DOI":"10.3389\/fnbot.2023.1096083","type":"journal-article","created":{"date-parts":[[2023,2,14]],"date-time":"2023-02-14T17:09:16Z","timestamp":1676394556000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":15,"title":["A multi-scale pooling convolutional neural network for accurate steel surface defects classification"],"prefix":"10.3389","volume":"17","author":[{"given":"Guizhong","family":"Fu","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Zengguang","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Wenwu","family":"Le","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jinbin","family":"Li","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Qixin","family":"Zhu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fuzhou","family":"Niu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Hao","family":"Chen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Fangyuan","family":"Sun","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yehu","family":"Shen","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2023,2,14]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"1447","DOI":"10.1109\/ICIEA.2012.6360951","article-title":"\u201cA fast method of steel surface defect detection using decision trees applied to lbp based features,\u201d","author":"Aghdam","year":"2012","journal-title":"2012 7th IEEE Conference on Industrial Electronics and Applications (ICIEA)"},{"key":"B2","doi-asserted-by":"publisher","first-page":"15849","DOI":"10.1073\/pnas.1903070116","article-title":"Reconciling modern machine-learning practice and the classical bias-variance trade-off","volume":"116","author":"Belkin","year":"2019","journal-title":"Proc. Nat. Acad. Sci. U.S.A"},{"key":"B3","doi-asserted-by":"publisher","first-page":"838","DOI":"10.1016\/j.jvcir.2013.05.011","article-title":"Automated defect detection in uniform and structured fabrics using gabor filters and pca","volume":"24","author":"Bissi","year":"2013","journal-title":"J. Vis. Commun. Image Represent"},{"key":"B4","doi-asserted-by":"publisher","first-page":"23","DOI":"10.1109\/SSD52085.2021.9429420","article-title":"\u201cDeep pre-trained models for computer vision applications: traffic sign recognition,\u201d","author":"Bouaafia","year":"2021","journal-title":"2021 18th International Multi-Conference on Systems, Signals and Devices (SSD)"},{"key":"B5","doi-asserted-by":"publisher","first-page":"140","DOI":"10.1016\/j.chemolab.2017.10.020","article-title":"Steel surface defects recognition based on multi-type statistical features and enhanced twin support vector machine","volume":"171","author":"Chu","year":"2017","journal-title":"Chemometr. Intellig. Lab. Syst"},{"key":"B6","doi-asserted-by":"publisher","first-page":"704","DOI":"10.1063\/1.1916744","article-title":"Defect detection using hidden markov random fields","volume":"760","author":"Dogand\u017ei\u0107","year":"2005","journal-title":"AIP Conf. Proc"},{"key":"B7","doi-asserted-by":"publisher","first-page":"397","DOI":"10.1016\/j.optlaseng.2019.05.005","article-title":"A deep-learning-based approach for fast and robust steel surface defects classification","volume":"121","author":"Fu","year":"2019","journal-title":"Opt. Lasers Eng"},{"key":"B8","first-page":"249","article-title":"\u201cUnderstanding the difficulty of training deep feedforward neural networks,\u201d","author":"Glorot","year":"2010","journal-title":"Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics"},{"key":"B9","doi-asserted-by":"publisher","first-page":"610","DOI":"10.1109\/TSMC.1973.4309314","article-title":"Textural features for image classification","volume":"6","author":"Haralick","year":"1973","journal-title":"IEEE Trans. Syst. Man Cybern"},{"key":"B10","doi-asserted-by":"publisher","first-page":"770","DOI":"10.1109\/CVPR.2016.90","article-title":"\u201cDeep residual learning for image recognition,\u201d","author":"He","year":"2016","journal-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition"},{"key":"B11","article-title":"Mobilenets: efficient convolutional neural networks for mobile vision applications","author":"Howard","year":"2017","journal-title":"arXiv preprint"},{"key":"B12","article-title":"Squeezenet: alexnet-level accuracy with 50x fewer parameters and < 0.5mb model size","author":"Iandola","year":"2016","journal-title":"arXiv preprint"},{"key":"B13","doi-asserted-by":"publisher","first-page":"675","DOI":"10.1145\/2647868.2654889","article-title":"\u201cCaffe: convolutional architecture for fast feature embedding,\u201d","author":"Jia","year":"2014","journal-title":"Proceedings of the 22nd ACM International Conference on Multimedia"},{"key":"B14","doi-asserted-by":"publisher","first-page":"221808","DOI":"10.1109\/ACCESS.2020.3041849","article-title":"A universal and adaptive fabric defect detection algorithm based on sparse dictionary learning","volume":"8","author":"Kang","year":"2020","journal-title":"IEEE Access"},{"key":"B15","doi-asserted-by":"crossref","first-page":"261","DOI":"10.1109\/ISITIA.2017.8124091","article-title":"\u201cWelding defect classification based on convolution neural network (cnn) and gaussian kernel,\u201d","volume-title":"2017 International Seminar on Intelligent Technology and Its Applications (ISITIA)","author":"Khumaidi","year":"2017"},{"key":"B16","first-page":"1097","article-title":"\u201cImagenet classification with deep convolutional neural networks,\u201d","author":"Krizhevsky","year":"2012","journal-title":"International Conference on Neural Information Processing Systems"},{"key":"B17","doi-asserted-by":"publisher","first-page":"2278","DOI":"10.1109\/5.726791","article-title":"Gradient-based learning applied to document recognition","volume":"86","author":"LeCun","year":"1998","journal-title":"Proc. IEEE"},{"key":"B18","doi-asserted-by":"publisher","first-page":"587","DOI":"10.1080\/00405000.2014.929790","article-title":"Fabric defect detection based on multi-scale wavelet transform and gaussian mixture model method","volume":"106","author":"Li","year":"2015","journal-title":"J. Textile Inst"},{"key":"B19","doi-asserted-by":"publisher","first-page":"742","DOI":"10.1016\/j.patcog.2011.07.025","article-title":"Wavelet-based defect detection in solar wafer images with inhomogeneous texture","volume":"45","author":"Li","year":"2012","journal-title":"Pattern Recognit"},{"key":"B20","doi-asserted-by":"publisher","DOI":"10.1002\/srin.201600068","article-title":"An end-to-end steel strip surface defects recognition system based on convolutional neural networks","author":"Li","year":"2016","journal-title":"Steel Res. Int"},{"key":"B21","article-title":"Network in network","author":"Lin","year":"2013","journal-title":"arXiv preprint"},{"key":"B22","doi-asserted-by":"publisher","first-page":"4629","DOI":"10.3390\/ma13204629","article-title":"A light-weight deep-learning model with multi-scale features for steel surface defect classification","volume":"13","author":"Liu","year":"2020","journal-title":"Materials"},{"key":"B23","doi-asserted-by":"publisher","first-page":"10437","DOI":"10.1109\/CVPR42600.2020.01045","article-title":"\u201c12-in-1: Multi-task vision and language representation learning,\u201d","author":"Lu","year":"2020","journal-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition"},{"key":"B24","doi-asserted-by":"publisher","first-page":"1585","DOI":"10.1016\/j.imavis.2009.03.007","article-title":"Fabric defect detection using morphological filters","volume":"27","author":"Mak","year":"2009","journal-title":"Image Vis. Comput"},{"key":"B25","doi-asserted-by":"publisher","DOI":"10.1109\/ICMLA.2012.19","article-title":"\u201cAdaptive extended local ternary pattern (aeltp) for recognizing avatar faces,\u201d","author":"Mohamed","year":"2013","journal-title":"International Conference on Machine Learning and Applications"},{"key":"B26","doi-asserted-by":"publisher","first-page":"240","DOI":"10.1117\/12.2524681","article-title":"Performance analysis of machine learning and deep learning architectures for malaria detection on cell images","volume":"11139","author":"Narayanan","year":"2019","journal-title":"Appl. Mach. Learn"},{"key":"B27","doi-asserted-by":"publisher","DOI":"10.1186\/1687-5281-2014-50","article-title":"Review of vision-based steel surface inspection systems","author":"Neogi","year":"2014","journal-title":"EURASIP J. Image Video Process"},{"key":"B28","doi-asserted-by":"publisher","first-page":"3260","DOI":"10.1016\/j.eswa.2010.09.012","article-title":"Machine learning approach for automated visual inspection of machine components","volume":"38","author":"Ravikumar","year":"2011","journal-title":"Expert Syst. Appl"},{"key":"B29","doi-asserted-by":"publisher","first-page":"929","DOI":"10.1109\/TCYB.2017.2668395","article-title":"A generic deep-learning-based approach for automated surface inspection","volume":"48","author":"Ren","year":"2018","journal-title":"IEEE Trans. Cybern"},{"key":"B30","author":"Simonyan","year":"2015","journal-title":"Very Deep Convolutional Networks for Large-Scale Image Recognition"},{"key":"B31","doi-asserted-by":"publisher","first-page":"858","DOI":"10.1016\/j.apsusc.2013.09.002","article-title":"A noise robust method based on completed local binary patterns for hot-rolled steel strip surface defects","volume":"285","author":"Song","year":"2013","journal-title":"Appl. Surf. Sci"},{"key":"B32","first-page":"2579","article-title":"Visualizing data using t-sne","volume":"9","author":"Van der Maaten","year":"2008","journal-title":"J. Mach. Learn. Res"},{"key":"B33","doi-asserted-by":"publisher","first-page":"517","DOI":"10.1109\/ICMLA.2018.00083","article-title":"\u201cAutomatic recognition of mild cognitive impairment and alzheimers disease using ensemble based 3d densely connected convolutional networks,\u201d","author":"Wang","year":"2018","journal-title":"2018 17th IEEE International Conference on Machine Learning and Applications (ICMLA)"},{"key":"B34","doi-asserted-by":"publisher","first-page":"426","DOI":"10.1109\/TSMC.2020.2997852","article-title":"An ensemble-based densely-connected deep learning system for assessment of skeletal maturity","volume":"52","author":"Wang","year":"2020","journal-title":"IEEE Trans. Syst. Man Cybern. Syst"},{"key":"B35","article-title":"How neural networks extrapolate: from feedforward to graph neural networks","author":"Xu","year":"2020","journal-title":"arXiv preprint"},{"key":"B36","doi-asserted-by":"publisher","first-page":"4945","DOI":"10.1109\/TNNLS.2021.3063516","article-title":"Tensorizing gan with high-order pooling for Alzheimer's disease assessment","volume":"33","author":"Yu","year":"2021","journal-title":"IEEE Trans. Neural Netw. Learn. Syst"},{"key":"B37","doi-asserted-by":"publisher","first-page":"51","DOI":"10.1016\/j.jmapro.2019.04.023","article-title":"Random forest-based real-time defect detection of al alloy in robotic arc welding using optical spectrum","volume":"42","author":"Zhang","year":"2019","journal-title":"J. Manuf. Process"},{"key":"B38","doi-asserted-by":"publisher","first-page":"2921","DOI":"10.1109\/CVPR.2016.319","article-title":"\u201cLearning deep features for discriminative localization,\u201d","author":"Zhou","year":"2016","journal-title":"Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition"},{"key":"B39","doi-asserted-by":"publisher","first-page":"400","DOI":"10.1016\/j.isatra.2021.06.025","article-title":"Classification of biscuit tiles for defect detection using fourier transform features","volume":"125","author":"Zori\u0107","year":"2022","journal-title":"ISA Trans"}],"container-title":["Frontiers in Neurorobotics"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnbot.2023.1096083\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,2,14]],"date-time":"2023-02-14T17:09:27Z","timestamp":1676394567000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fnbot.2023.1096083\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,2,14]]},"references-count":39,"alternative-id":["10.3389\/fnbot.2023.1096083"],"URL":"https:\/\/doi.org\/10.3389\/fnbot.2023.1096083","relation":{},"ISSN":["1662-5218"],"issn-type":[{"value":"1662-5218","type":"electronic"}],"subject":[],"published":{"date-parts":[[2023,2,14]]},"article-number":"1096083"}}