125361 20251113150203.0 doi 10.1109/TII.2023.3234030 sideral 133106 ART-2023-133106 spa Sampath, V. Attention guided multi-task learning for surface defect identification 2023 Access copy available to the general public Unrestricted Surface defect identification is an essential task in the industrial quality control process, in which visual checks are conducted on a manufactured product to ensure that it meets quality standards. Convolutional Neural Network (CNN) based surface defect identification method has proven to outperform traditional image processing techniques. However, the real-world surface defect datasets are limited in size due to the expensive data generation process and the rare occurrence of defects. To address this issue, this paper presents a method for exploiting auxiliary information beyond the primary labels to improve the generalization ability of surface defect identification tasks. Considering the correlation between pixel level segmentation masks, object level bounding boxes and global image level classification labels, we argue that jointly learning features of the related tasks can improve the performance of surface defect identification tasks. This paper proposes a framework named Defect-Aux-Net, based on multi-task learning with attention mechanisms that exploit the rich additional information from related tasks with the goal of simultaneously improving robustness and accuracy of the CNN based surface defect identification. We conducted a series of experiments with the proposed framework. The experimental results showed that the proposed method can significantly improve the performance of state-of-the-art models while achieving an overall accuracy of 97.1%, Dice score of 0.926 and mAP of 0.762 on defect classification, segmentation and detection tasks. info:eu-repo/grantAgreement/EC/H2020/814225/EU/DIGItal MANufacturing Technologies for Zero-defect Industry 4.0 Production/DIGIMAN4.0 This project has received funding from the European Union’s Horizon 2020 research and innovation program under grant agreement No H2020 814225-DIGIMAN4.0 info:eu-repo/semantics/openAccess by https://creativecommons.org/licenses/by/4.0/deed.es 11.7 2023 AUTOMATION & CONTROL SYSTEMS 2 / 84 = 0.024 2023 Q1 T1 ENGINEERING, INDUSTRIAL 2 / 69 = 0.029 2023 Q1 T1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS 3 / 170 = 0.018 2023 Q1 T1 4.42 2023 Computer Science Applications 2023 Q1 Information Systems 2023 Q1 Electrical and Electronic Engineering 2023 Q1 Control and Systems Engineering 2023 Q1 24.1 2023 info:eu-repo/semantics/article info:eu-repo/semantics/publishedVersion Maurtua, I. Aguilar Martin, J. J. Universidad de Zaragoza (orcid)0000-0002-8609-1358 Rivera, A. Molina, J. Gutiérrez, A. 5002 515 Universidad de Zaragoza Dpto. Ingeniería Diseño Fabri. Área Ing. Procesos Fabricación 19, 9 (2023), 9713-9721 IEEE Trans. Ind. Inform. IEEE Transactions on Industrial Informatics 1551-3203 18338858 http://zaguan.unizar.es/record/125361/files/texto_completo.pdf Versión publicada 3541079 http://zaguan.unizar.es/record/125361/files/texto_completo.jpg?subformat=icon icon Versión publicada oai:zaguan.unizar.es:125361 articulos driver 2025-11-13-15:00:44 ARTICLE