{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,17]],"date-time":"2026-05-17T09:09:30Z","timestamp":1779008970511,"version":"3.51.4"},"reference-count":40,"publisher":"SAGE Publications","issue":"6","license":[{"start":{"date-parts":[[2025,11,10]],"date-time":"2025-11-10T00:00:00Z","timestamp":1762732800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/journals.sagepub.com\/page\/policies\/text-and-data-mining-license"}],"content-domain":{"domain":["journals.sagepub.com"],"crossmark-restriction":true},"short-container-title":["Journal of Intelligent &amp; Fuzzy Systems: Applications in Engineering and Technology"],"published-print":{"date-parts":[[2026,6]]},"abstract":"<jats:title>Abstracts<\/jats:title>\n                  <jats:p>Automotive underbody sealant defect detection is a critical task for quality control. However, conventional semantic segmentation models often struggle with challenges such as low accuracy on small defects, imprecise edge segmentation, and high computational costs. To address these issues, this paper proposes ACS-DeepLabv3+, a lightweight and accurate network. The model enhances efficiency by replacing the original backbone with MobileNetV2 and introduces a novel, empirically validated Car-ASPP module. This module improves multi-scale feature extraction by integrating depthwise separable convolutions with dual attention mechanisms: the Convolutional Block Attention Module (CBAM) and the Selective Kernel Network (SKNet). A transfer learning strategy with early stopping is also employed to optimize the training process. On our industrial dataset, which features a variety of challenging defect types, ACS-DeepLabv3\u2009+\u2009achieves a mean Intersection over Union (mIoU) of 85.99% and a mean Pixel Accuracy (mPA) of 91.55%. With only 6.85\u2005M parameters and an inference speed of 33.52 FPS, our model significantly outperforms the original DeepLabv3\u2009+\u2009and other mainstream networks in both segmentation accuracy and computational efficiency, offering a robust solution for real-time industrial applications.<\/jats:p>","DOI":"10.1177\/18758967251390733","type":"journal-article","created":{"date-parts":[[2025,11,10]],"date-time":"2025-11-10T17:16:47Z","timestamp":1762795007000},"page":"1824-1840","update-policy":"https:\/\/doi.org\/10.1177\/sage-journals-update-policy","source":"Crossref","is-referenced-by-count":0,"title":["ACS-DeepLabv3+: A Deep Learning Network for Defect Detection in Automotive Manufacturing"],"prefix":"10.1177","volume":"50","author":[{"ORCID":"https:\/\/orcid.org\/0000-0001-8458-8398","authenticated-orcid":false,"given":"Chilan","family":"Cai","sequence":"first","affiliation":[{"name":"School of Intelligent Manufacturing and Control\u00a0Engineering, Shanghai Polytechnic University, Shanghai, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-4425-5662","authenticated-orcid":false,"given":"Zhigang","family":"Zhou","sequence":"additional","affiliation":[{"name":"School of Intelligent Manufacturing and Control\u00a0Engineering, Shanghai Polytechnic University, Shanghai, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0005-9170-6723","authenticated-orcid":false,"given":"Jiale","family":"Zhang","sequence":"additional","affiliation":[{"name":"School of Intelligent Manufacturing and Control\u00a0Engineering, Shanghai Polytechnic University, Shanghai, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0009-0009-5187-1942","authenticated-orcid":false,"given":"Jinwen","family":"Yu","sequence":"additional","affiliation":[{"name":"School of Intelligent Manufacturing and Control\u00a0Engineering, Shanghai Polytechnic University, Shanghai, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"179","published-online":{"date-parts":[[2025,11,10]]},"reference":[{"key":"e_1_3_2_2_1","volume-title":"The true cost of manual inspection in manufacturing [Internet]","author":"Akridata Inc","year":"2024","unstructured":"Akridata Inc. (2024). The true cost of manual inspection in manufacturing [Internet]. Akridata Inc. [cited 2025 Aug 20]. Available from: https:\/\/akridata.com\/the-true-cost-of-manual-inspection-in-manufacturing\/."},{"issue":"2","key":"e_1_3_2_3_1","first-page":"45","article-title":"Farmers\u2019 traditional management of trees outside forests in Bangladesh","volume":"16","author":"Alam M.","year":"2011","unstructured":"Alam M., Furukawa Y. (2011). Farmers\u2019 traditional management of trees outside forests in Bangladesh. J For Plann, 16(2), 45\u201351. https:\/\/doi.org\/10.20659\/jfp.16.2_45","journal-title":"J For Plann"},{"key":"e_1_3_2_4_1","doi-asserted-by":"publisher","DOI":"10.3390\/metrology4020016"},{"key":"e_1_3_2_5_1","doi-asserted-by":"publisher","DOI":"10.1016\/j.eswa.2024.123359"},{"key":"e_1_3_2_6_1","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2017.2699184"},{"key":"e_1_3_2_7_1","volume-title":"Encoder-decoder with atrous separable convolution for semantic image segmentation2018 Sep 8\u201314; Munich, Germany. Cham: Springer; pp. 833\u2013851.","author":"Chen L. C.","year":"2018","unstructured":"Chen L. C., Zhu Y., Papandreou G., et\u00a0al. (2018). Encoder-decoder with atrous separable convolution for semantic image segmentation. Proceedings of the European conference on computer vision (ECCV); 2018 Sep 8\u201314; Munich, Germany. Cham: Springer; pp. 833\u2013851. https:\/\/doi.org\/10.1007\/978-3-030-01234-2_49"},{"key":"e_1_3_2_8_1","volume-title":"Xception: Deep learning with depthwise separable convolutions2017 Jul 21\u201326; Honolulu, HI. New York: IEEE. pp. 1800\u20131807.","author":"Chollet F.","year":"2017","unstructured":"Chollet F. (2017). Xception: Deep learning with depthwise separable convolutions. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR); 2017 Jul 21\u201326; Honolulu, HI. New York: IEEE. pp. 1800\u20131807. https:\/\/doi.org\/10.1109\/CVPR.2017.195"},{"key":"e_1_3_2_9_1","volume-title":"What is a good evaluation measure for semantic segmentation?Proceedings of the British Machine Vision Conference (BMVC)","author":"Csurka G.","year":"2013","unstructured":"Csurka G., Larlus D., Perronnin F. (2013). What is a good evaluation measure for semantic segmentation? Proceedings of the British Machine Vision Conference (BMVC); 2013 Sep 9\u201313; Bristol, UK. Durham: BMVA Press. pp. 32.1\u201312. https:\/\/doi.org\/10.5244\/C.27.32"},{"key":"e_1_3_2_10_1","volume-title":"ImageNet: A large-scale hierarchical image database","author":"Deng J.","year":"2009","unstructured":"Deng J., Dong W., Socher R., et\u00a0al. (2009). ImageNet: A large-scale hierarchical image database. 2009 IEEE Conference on Computer Vision and Pattern Recognition; 2009 Jun 20\u201325; Miami, FL. New York: IEEE. pp. 248\u2013255. https:\/\/doi.org\/10.1109\/CVPR.2009.5206848"},{"key":"e_1_3_2_11_1","unstructured":"Gholamy A. Kreinovich V. Kosheleva O. (2018) Why 70\/30 or 80\/20 relation between training and testing sets: a pedagogical explanation. El Paso (TX): University of Texas at El Paso Department of Computer Science; Report No.: UTEP-CS-18-09. Available from: https:\/\/scholarworks.utep.edu\/cs_techrep\/1209."},{"issue":"1","key":"e_1_3_2_12_1","first-page":"45","article-title":"Enhancing lung cancer detection: Optimizing deep learning with convolutional block attention module","volume":"5","author":"Gokavarapu S.","year":"2025","unstructured":"Gokavarapu S., Rao K. V., Srinivas G., et\u00a0al. (2025). Enhancing lung cancer detection: Optimizing deep learning with convolutional block attention module. J Artif Intell Technol, 5(1), 45\u201360. https:\/\/doi.org\/10.37965\/jait.2025.0668","journal-title":"J Artif Intell Technol"},{"key":"e_1_3_2_13_1","unstructured":"Hao Y. Liu Y. Chen Y. Han L. Peng J. Tang S. et\u00a0al. (2022) EISeg: an efficient interactive segmentation tool based on PaddlePaddle. arXiv [Preprint]. [cited 2025 Aug 20]. Available from: http:\/\/arxiv.org\/abs\/2210.08788. https:\/\/doi.org\/10.48550\/arXiv.2210.08788"},{"key":"e_1_3_2_14_1","unstructured":"Howard A. G. Zhu M. Chen B. et\u00a0al. (2017) MobileNets: Efficient convolutional neural networks for mobile vision applications. arXiv [Preprint]. [cited 2025 Aug 20]. Available from: http:\/\/arxiv.org\/abs\/1704.04861."},{"key":"e_1_3_2_15_1","doi-asserted-by":"publisher","DOI":"10.3390\/polym13040525"},{"key":"e_1_3_2_16_1","doi-asserted-by":"publisher","DOI":"10.3390\/agriculture12081126"},{"issue":"1","key":"e_1_3_2_17_1","first-page":"679","article-title":"Enhancing tea leaf disease identification with lightweight MobileNetV2","volume":"80","author":"Li Z.","unstructured":"Li Z., Li Y., Yan C., et\u00a0al. (2024a). Enhancing tea leaf disease identification with lightweight MobileNetV2. Comput Mater Contin, 80(1), 679\u2013694. https:\/\/doi.org\/10.32604\/cmc.2024.051526","journal-title":"Comput Mater Contin"},{"key":"e_1_3_2_18_1","doi-asserted-by":"publisher","DOI":"10.3390\/s24237691"},{"key":"e_1_3_2_19_1","unstructured":"Loshchilov I. Hutter F. (2016) [cited 2025 Aug 20]. SGDR: stochastic gradient descent with warm restarts. arXiv [Preprint]. Available from: http:\/\/arxiv.org\/abs\/1608.03983. https:\/\/doi.org\/10.48550\/arXiv.1608.03983"},{"key":"e_1_3_2_20_1","unstructured":"Loshchilov I. Hutter F. (2017) [cited 2025 Aug 20]. Decoupled weight decay regularization. arXiv [Preprint]. Available from: http:\/\/arxiv.org\/abs\/1711.05101. https:\/\/doi.org\/10.48550\/arXiv.1711.05101"},{"issue":"7","key":"e_1_3_2_21_1","first-page":"3523","article-title":"Image segmentation using deep learning: A survey","volume":"44","author":"Minaee S.","year":"2022","unstructured":"Minaee S., Boykov Y., Porikli F., et\u00a0al. (2022). Image segmentation using deep learning: A survey. IEEE Trans Pattern Anal Mach Intell, 44(7), 3523\u20133542. https:\/\/doi.org\/10.1109\/TPAMI.2021.3059968","journal-title":"IEEE Trans Pattern Anal Mach Intell"},{"issue":"4","key":"e_1_3_2_22_1","first-page":"7123","article-title":"A systematic review for quality within the automotive industry 4.0","volume":"20","author":"Popescu D.","year":"2025","unstructured":"Popescu D., Dragomir M., Popescu S., et\u00a0al. (2025). A systematic review for quality within the automotive industry 4.0. Int J Comput Commun Control, 20(4), 7123. https:\/\/doi.org\/10.15837\/ijccc.2025.4.7123","journal-title":"Int J Comput Commun Control"},{"key":"e_1_3_2_23_1","doi-asserted-by":"publisher","DOI":"10.1109\/ACCESS.2024.3351805"},{"key":"e_1_3_2_24_1","unstructured":"Rahman A. Seethi V. D. R. Yunker A. Kral Z. Kettimuthu R. Alhoori H. (2025). Effective defect detection using instance segmentation for NDI. arXiv [Preprint]. [cited 2025 Aug 20]. Available from: http:\/\/arxiv.org\/abs\/2501.14149. https:\/\/doi.org\/10.48550\/arXiv.2501.14149"},{"key":"e_1_3_2_25_1","volume-title":"U-Net:convolutional networks for biomedical image segmentationMedical Image Computing and Computer-Assisted Intervention \u2013 MICCAI 2015","author":"Ronneberger O.","year":"2015","unstructured":"Ronneberger O., Fischer P., Brox T. (2015). U-Net:convolutional networks for biomedical image segmentation. Medical Image Computing and Computer-Assisted Intervention \u2013 MICCAI 2015; 2015 Oct 5-9; Munich, Germany. Cham: Springer; pp. 234\u2013241. https:\/\/doi.org\/10.1007\/978-3-319-24574-4_28"},{"key":"e_1_3_2_26_1","doi-asserted-by":"publisher","DOI":"10.1201\/9781003322641-34"},{"key":"e_1_3_2_27_1","volume-title":"MobileNetV2: Inverted residuals and linear bottlenecks","author":"Sandler M.","year":"2018","unstructured":"Sandler M., Howard A., Zhu M., et\u00a0al. (2018). MobileNetV2: Inverted residuals and linear bottlenecks. 2018 IEEE\/CVF Conference on Computer Vision and Pattern Recognition; 2018 Jun 18\u201323; Salt Lake City, UT. New York: IEEE; pp. 4510\u20134520. https:\/\/doi.org\/10.1109\/CVPR.2018.00474"},{"key":"e_1_3_2_28_1","doi-asserted-by":"publisher","DOI":"10.1186\/s40537-019-0197-0"},{"key":"e_1_3_2_29_1","first-page":"107254","article-title":"An enhanced ResNet deep learning method for multimodal signal-based locomotion intention recognition","volume":"98","author":"Sun H.","unstructured":"Sun H., Gu X., Zhang Y., Sun F., Zhang S., Wang D., et\u00a0al. (2024a). An enhanced ResNet deep learning method for multimodal signal-based locomotion intention recognition. Biomed Signal Process Control, 98, 107254. https:\/\/doi.org\/10.1016\/j.bspc.2024.107254","journal-title":"Biomed Signal Process Control"},{"key":"e_1_3_2_30_1","doi-asserted-by":"publisher","DOI":"10.1109\/AIPMV62663.2024.10692274"},{"key":"e_1_3_2_31_1","doi-asserted-by":"publisher","DOI":"10.1007\/s12145-025-01793-1"},{"key":"e_1_3_2_32_1","unstructured":"Thoma M. (2016) A survey of semantic segmentation. arXiv [Preprint]. [cited 2025 Aug 20]. Available from: http:\/\/arxiv.org\/abs\/1602.06541. https:\/\/doi.org\/10.48550\/arXiv.1602.06541"},{"key":"#cr-split#-e_1_3_2_33_1.1","unstructured":"Tulip Interfaces. (2024) Automated visual inspection (AVI): a comprehensive guide [Internet]. Somerville (MA): Tulip Interfaces Inc."},{"key":"#cr-split#-e_1_3_2_33_1.2","unstructured":"[cited 2025 Aug 20]. Available from: https:\/\/www.matroid.com\/visual-inspection-a-comprehensive-guide\/."},{"key":"e_1_3_2_34_1","doi-asserted-by":"publisher","DOI":"10.3390\/s25072043"},{"key":"e_1_3_2_35_1","doi-asserted-by":"publisher","DOI":"10.1186\/s12880-025-01625-0"},{"key":"e_1_3_2_36_1","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2020.2983686"},{"key":"e_1_3_2_37_1","unstructured":"Zhang Z. Niu C. Zhao Z. Zhang X. Chen X. (2024) Small object few-shot segmentation for vision-based industrial inspection. arXiv [Preprint]. [cited 2025 Aug 20]. Available from: http:\/\/arxiv.org\/abs\/2407.21351. https:\/\/doi.org\/10.48550\/arXiv.2407.21351"},{"key":"e_1_3_2_38_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.660"},{"issue":"2","key":"e_1_3_2_39_1","first-page":"1564","article-title":"Multi-class segmentation of navel orange surface defects based on improved DeepLabv3+","volume":"55","author":"Zhu Y.","year":"2024","unstructured":"Zhu Y., Liu S., Wu X., Gao L., Xu Y. (2024). Multi-class segmentation of navel orange surface defects based on improved DeepLabv3+. J Agric Eng, 55(2), 1564. https:\/\/doi.org\/10.4081\/jae.2024.1564","journal-title":"J Agric Eng"},{"key":"e_1_3_2_40_1","doi-asserted-by":"publisher","DOI":"10.1109\/JPROC.2020.3004555"}],"container-title":["Journal of Intelligent &amp; Fuzzy Systems: Applications in Engineering and Technology"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/journals.sagepub.com\/doi\/pdf\/10.1177\/18758967251390733","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.sagepub.com\/doi\/full-xml\/10.1177\/18758967251390733","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.sagepub.com\/doi\/pdf\/10.1177\/18758967251390733","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,17]],"date-time":"2026-05-17T08:32:34Z","timestamp":1779006754000},"score":1,"resource":{"primary":{"URL":"https:\/\/journals.sagepub.com\/doi\/10.1177\/18758967251390733"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,11,10]]},"references-count":40,"journal-issue":{"issue":"6","published-print":{"date-parts":[[2026,6]]}},"alternative-id":["10.1177\/18758967251390733"],"URL":"https:\/\/doi.org\/10.1177\/18758967251390733","relation":{},"ISSN":["1064-1246","1875-8967"],"issn-type":[{"value":"1064-1246","type":"print"},{"value":"1875-8967","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,11,10]]}}}