Abstract
Few-shot defect detection holds significant importance for adapting to complex industrial environments and enhancing detection accuracy. Addressing the issue where existing few-shot defect detection methods are prone to compromised feature representation under varying defect scales, this study proposes a distance-guided prototype network (DGPN) with explicit meta learning. Building upon the prototype network, our method leverages distance information between normal and anomalous image features to guide feature transformation and fusion processes. An attention gating block (AGB) is introduced to convert distance representation vectors into feature maps while enhancing the capture capability for fine-grained targets. A multiscale feature fusion module (MSFF) is proposed to further strengthen the network’s ability to extract multiscale features and semantic information. Additionally, an upsample fusion module (UF) is designed to fully integrate and exchange multiscale contextual information between shallow and deep layers, decoding semantic information and spatial details to improve the precision of small defect detection. Extensive experiments on Industrial-5i, Visa, and a self-built chip dataset validate the superiority of the proposed method and the practical applicability.
This is a preview of subscription content, log in via an institution to check access.
Data availability
Data will be made available on request.
References
Chang, Z., Lu, Y., Ran, X., Gao, X., & Wang, X. (2023). Few-shot semantic segmentation: A review on recent approaches. Neural Computing and Applications, 35, 18251–18275.
Chen, H., Yu, Y., Dong, Y., Lu, Z., Li, Y., & Zhang, Z. (2024). Multi-content interaction network for few-shot segmentation. ACM Transactions on Multimedia Computing, Communications, and Applications, 20, 177.
Finn C., Abbeel P., & Levine S. (2017) Model-agnostic meta-learning for fast adaptation of deep networks. In International Conference on Machine Learning (ICML), PMLR, 2017, pp. 1126–1135.
Gong, L., Zhang, F., Zhao, Y., & Duan, J. (2026). Frequency-aware and global-local selective attention network for laser welding spot detection. Neurocomputing, 678, Article 122187.
Gong, Y., Wang, X., Zhou, C., Ge, M., Liu, C., & Zhang, X. (2025). Human-machine knowledge hybrid augmentation method for surface defect detection based few-data learning. Journal of Intelligent Manufacturing, 36, 1723–1742.
Ha, T., Hwang, C., & Jeong, J. (2026). CLIP-MDC: CLIP encoder based multimodal defect classification with synthetic anomaly generation for real-time surface defect detection. Journal of Intelligent Manufacturing. https://doi.org/10.1007/s10845-025-02773-4
Hou, Y., Fan, H., Chen, Y., & Liu, G. (2025). A chip inspection system based on a multiscale subarea attention network. Journal of Intelligent Manufacturing, 36, 4039–4053.
Hu, W., Zhang, F., Zhao, Y., & Duan, J. (2026). Physics knowledge-inspired scattering neural representation for micro-adhesive-spot segmentation under complex backgrounds. IEEE Transactions on Industrial Informatics, 22(6), 5287.
Huang, Q., Zhang, F., Zhao, Y., & Duan, J. (2025). Frequency-domain multi-scale Kolmogorov-Arnold representation attention network for mixed-type wafer defect recognition. Engineering Applications of Artificial Intelligence, 144, Article 110121.
Li X., Wei T., Chen Y. P., Tai Y.W., & Tang C.K. (2020) FSS-1000: A 1000-Class dataset for few-shot segmentation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 2869–2878.
Li G., Jampani V., Lara L. S., Sun D., Kim J., & Kim J. (2021) Adaptive prototype learning and allocation for few-shot segmentation. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 8334–8343.
Li C. L., Sohn K., Yoon J., & Pfister T. (2021) Cutpaste: Self-supervised learning for anomaly detection and localization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 9664–9674.
Li, G., Peng, F., Wu, Z., Wang, S., & Xu, R. Y. D. (2024). ODCL: An object disentanglement and contrastive learning model for few-shot industrial defect detection. IEEE Sensors Journal, 24(11), 18568–18577.
Li, X., Fan, Z., Liu, Q., & Wan, X. (2025). DSP-YOLO: An improved YOLO11-based method for steel surface defect detection. Measurement Science and Technology, 36(9), Article 096201.
Liu, K., Yang, Y., Yang, X., Wang, J., Liu, W., & Chen, H. (2025a). Multi-level joint distributed alignment-based domain adaptation for cross-scenario strip defect recognition. Journal of Intelligent Manufacturing, 36, 2373–2386.
Liu, T., Li, B., Jin, X., Shi, Y., Li, Q., & Wei, X. (2025b). Exploring few-shot defect segmentation in general industrial scenarios with metric learning and vision foundation models. Optics and Laser Technology, 192, Article 114078.
Lu Y., Yu F., Reddy M., & Wang Y.(2020) Few-shot scene-adaptive anomaly detection. In Computer Vision-ECCV 2020: 16th European Conference, Glasgow, UK, 2020, pp. 125–141.
Ma, Y., Yin, J., Huang, F., & Li, Q. (2024). Surface defect inspection of industrial products with object detection deep networks: A systematic review. Artificial Intelligence Review, 57, 333.
Min J., Kang D., & Cho M.(2021) Hypercorrelation squeeze for few-shot segmentation. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV).
Shaban A., Bansal S., Liu Z., Essa I., & Boots B.(2017) One-shot learning for semantic segmentation. arXiv preprint arXiv:1709.03410, 2017.
Shi, X., Cui, Z., Zhang, S., Cheng, M., He, L., & Tang, X. (2022). Multi-similarity based hyperrelation network for few-shot segmentation. IET Image Processing, 17, 204–214.
Shi, X., Zhang, S., Cheng, M., He, L., Tang, X., & Cui, Z. (2023). Few-shot semantic segmentation for industrial defect recognition. Computers in Industry, 148, Article 103901.
Song, K., Feng, H., Cao, T., Cui, W., & Yan, Y. (2024). MFANet: Multifeature aggregation network for cross-granularity few-shot seamless steel tubes surface defect segmentation. IEEE Transactions on Industrial Informatics, 20(7), 9725–9735.
Sung F., Yang Y., Zhang L., Xiang T., Torr P. H. S., & Hospedales T. M. (2018) Learning to compare: relation network for few-shot learning. In Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. (CVPR), pp. 1199–1208.
Tang, C., Yi, J., Chen, Y., Wu, X., & Xiong, W. (2025). Bidirectional consistent hypercorrelation network for cross-domain few-shot segmentation. Knowledge-Based Systems, 331, Article 114866.
Wang K., Liew J. H., Zou Y., Zhou D., & Feng J. (2019) Panet: Few-shot image semantic segmentation with prototype alignment. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pp. 9197–9206.
Wang, H., Li, Z., & Wang, H. (2022). Few-shot steel surface defect detection. IEEE Transactions on Instrumentation and Measurement. https://doi.org/10.1109/tim.2021.3128208
Wang, S., Chen, H., Liu, K., Zhou, Y., & Feng, H. (2023). Meta-FSDet: A meta-learning based detector for few-shot defects of photovoltaic modules. Journal of Intelligent Manufacturing, 34, 3413–3427.
Wang, Y., Khan, W. A., & Chung, S. H. (2024). Few-shot defect detection of catheter products via enlarged scale feature pyramid and contrastive proposal memory bank. IEEE Transactions on Industrial Informatics, 20(11), 13036–13046.
Wu Z., Shi X., Lin G., & Cai J. (2021) Learning meta-class memory for few-shot semantic segmentation. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pp. 517–526.
Xiao, X., Zhao, Y., Zhang, F., Luo, B., Yu, L., Chen, B., & Yang, C. (2023). BASeg: Boundary aware semantic segmentation for autonomous driving. Neural Networks, 157, 460–470.
Xue, J., Gao, B., Liu, G., Zhang, Y., Woo, W. L., Yang, Y., & Yu, Y. (2023). Memory linked knowledge domain transfer few-shot learning for thermography nondestructive evaluation system. Infrared Physics & Technology, 132, Article 104718.
Yang B., Liu C., Zhang L., & Li X. (2020) Prototype mixture models for few-shot semantic segmentation. In Proceedings of the European Conference on Computer Vision (ECCV)
Zhang H., Wu Z., Wang Z., Chen Z., & Jiang Y.(2023) Prototypical residual networks for anomaly detection and localization. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pp. 16281–16291.
Zhang, F., Gong, L., Lin, S., & Zhao, Y. (2026b). Memory guided dual discrepancy network with domain adaptation for few-shot anomaly detection. Expert Systems with Applications, 314, Article 131711.
Zhang, F., Hu, H., Zou, B., & Luo, M. (2025b). M4Net: Multi-level multi-patch multi-receptive multi-dimensional attention network for infrared small target detection. Neural Networks, 183, Article 107026.
Zhang, F., Hu, W., Wang, Y., & Duan, J. (2026a). Dynamic background-guided asymmetric knowledge distillation network for 3D defect detection. Journal of Manufacturing Processes, 160, 185–199.
Zhang, F., Lv, Q., Pan, B., & Wang, Y. (2025a). Boundary semantic interactive aggregation network for scene segmentation. Expert Systems with Applications, 272, Article 126754.
Zhang, F., Pan, B., Wang, Y., Zhou, C., Zhao, Y., & Duan, J. (2026c). High-frequency semantic hierarchical dynamic feature fusion network for chip surface defect detection. Measurement Science and Technology, 37, Article 125401.
Zhong, S., Zhang, F., & Duan, J. (2023). Context-guided reverse attention network with multiscale aggregation for infrared small target detection. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 16, 9725.
Zhong, S., Zhou, H., Yan, Y., Zhang, F., & Duan, J. (2024). Asymmetric convolutional multi-level attention network for micro-lens segmentation. Engineering Applications of Artificial Intelligence, 133, Article 108355.
Zhou, X., Zhang, Y., Liu, Z., Jiang, Z., Ren, Z., Mi, T., & Zhou, S. (2025). IFIFusion: An independent feature information fusion model for surface defect detection. Information Fusion, 120, Article 103039.
Zou T., Xiong S., Yao R., & Rong Y. (2025) Balancing conservatism and aggressiveness: prototype-affinity hybrid network for few-shot segmentation. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), pp. 20561–20571.
Acknowledgements
This research was funded by National Key Research and Development Program of China (Grant No. 2024YFF0504904), Key Research and Development Plan Program of Hunan Province (Grant No. 2023GK2021), and the Project of State Key Laboratory of Precision Manufacturing for Extreme Service Performance of Central South University (Grant No. ZZYJKT2024-09).
Funding
National Key R & D Program of China,2024YFF0504904, Ji’an Duan, Key R & D Plan Program of Hunan Province, 2023GK2021, Yuqian Zhao, Project of State Key Laboratory of Precision Manufacturing for Extreme Service Performance of Central South University,ZZYJKT2024-09, Fan Zhang.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors indicated that they have no conflict of interest with this work.
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Zhang, F., Gong, L., Zhao, Y. et al. Distance-guided prototype network for few-shot vision-based industrial defect detection. J Intell Manuf (2026). https://doi.org/10.1007/s10845-026-02960-x
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1007/s10845-026-02960-x