Downloads

https://doi.org/10.62836/iaet.v5i1.0008
Liang, S., & Xiao, J. (2026). Deep Learning for Industrial Visual Inspection: A Review of Small-Target Detection, Lighting Robustness, and Imbalance-Aware Learning. Innovations in Applied Engineering and Technology, 5(1), 0008. https://doi.org/10.62836/iaet.v5i1.0008
Volume 5, Issue 1, 2026

Deep Learning for Industrial Visual Inspection: A Review of Small-Target Detection, Lighting Robustness, and Imbalance-Aware Learning

Deep learning has substantially increased the automation and accuracy of industrial visual inspection, but performance obtained under curated laboratory conditions often deteriorates once models are deployed on real production lines. The difficulty is especially pronounced when defects occupy only a few pixels, exhibit low contrast, are distorted by uneven illumination, or occur too rarely to support balanced training. Rather than treating these issues separately, this review examines their interaction through 50 representative studies. For small-target inspection, the discussion covers multi-scale pyramids, high-resolution prediction branches, cross-level feature exchange, and detail-preserving designs. For illumination robustness, it reviews image enhancement, illumination normalization, and attention-based feature calibration. For imbalanced learning, it compares resampling, synthetic defect generation, class-aware weighting, logit correction, and gradient rebalancing. Evidence from steel, printed circuit boards, textiles, concrete, and tunnel-lining inspection is used to contrast the strengths and limitations of these approaches. The review also highlights inconsistent definitions of small objects, limited control of lighting conditions, incomplete reporting for minority classes, and insufficient cross-device validation. A more standardized evaluation protocol is therefore advocated, combining scale-, illumination-, and class-distribution-specific metrics with latency and computational-cost reporting.

industrial visual inspection small-target detection illumination robustness imbalanced learning deep learning intelligent manufacturing

References

  1. Czimmermann T, Ciuti G, Milazzo M, et al. Visual-Based Defect Detection and Classification Approaches for Industrial Applications—A Survey. Sensors 2020; 20: 1459. https://doi.org/10.3390/s20051459.
  2. Yang J, Li S, Wang Z, et al. Using Deep Learning to Detect Defects in Manufacturing: A Comprehensive Survey and Current Challenges. Materials 2020; 13: 5755. https://doi.org/10.3390/ma13245755.
  3. Gao Y, Li X, Wang XV, et al. A Review on Recent Advances in Vision-Based Defect Recognition towards Industrial Intelligence. Journal of Manufacturing Systems 2022; 62: 753–766. https://doi.org/10.1016/j.jmsy.2021.05.008.
  4. Jha SB, Babiceanu RF. Deep CNN-Based Visual Defect Detection: Survey of Current Literature. Computers in Industry 2023; 148: 103911. https://doi.org/10.1016/j.compind.2023.103911.
  5. Liu Y, Zhang C, Dong X. A Survey of Real-Time Surface Defect Inspection Methods Based on Deep Learning. Artificial Intelligence Review 2023; 56: 12131–12170. https://doi.org/10.1007/s10462-023-10475-7.
  6. Ma Y, Yin J, Huang F, et al. Surface Defect Inspection of Industrial Products with Object Detection Deep Networks: A Systematic Review. Artificial Intelligence Review 2024; 57: 333. https://doi.org/10.1007/s10462-024-10956-3.
  7. Ren S, He K, Girshick R, et al. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks. IEEE Transactions on Pattern Analysis and Machine Intelligence 2017; 39: 1137–1149. https://doi.org/10.1109/tpami.2016.2577031.
  8. Redmon J, Divvala S, Girshick R, et al. You Only Look Once: Unified, Real-Time Object Detection. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA, 27–30 June 2016. https://doi.org/10.1109/cvpr.2016.91.
  9. Lin TY, Dollar P, Girshick R, et al. Feature Pyramid Networks for Object Detection. In Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA, 21–26 July 2017. https://doi.org/10.1109/cvpr.2017.106.
  10. Liu S, Qi L, Qin H, et al. Path Aggregation Network for Instance Segmentation. In Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA, 18–23 June 2018. https://doi.org/10.1109/cvpr.2018.00913.
  11. Tan M, Pang R, Le QV. EfficientDet: Scalable and Efficient Object Detection. In Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 13–19 June 2020. https://doi.org/10.1109/cvpr42600.2020.01079.
  12. Huang S, Lu Z, Cheng R, et al. FaPN: Feature-Aligned Pyramid Network for Dense Image Prediction. In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), Montreal, QC, Canada, 10–17 October 2021. https://doi.org/10.1109/iccv48922.2021.00090.
  13. Zhang Y, Sun Y, Huang Q, et al. Cross-Level Multiscale Feature Fusion Enhanced YOLO with Focused-CIoU for Detection of Small and Extreme Aspect Ratio Defects. Measurement Science and Technology 2026; 37: 025401. https://doi.org/10.1088/1361-6501/ae26a5.
  14. Ren J, Li Z, Zhou Z, et al. CFL-YOLO: A Modified YOLOv8 for Detection of Steel Surface Defects. Measurement 2026; 266: 120509. https://doi.org/10.1016/j.measurement.2026.120509.
  15. Li G, Gan Y, Zhang W, et al. GS-YOLO: A Lightweight and High-Performance Method for PCB Surface Defect Detection. Expert Systems with Applications 2026; 303: 130583. https://doi.org/10.1016/j.eswa.2025.130583.
  16. Li J, He X, Chen X, et al. HDFA-YOLO: A Real-Time Steel Surface Defect Detection Model Based on Backbone Lightweight Design and Multi-Scale Feature Fusion. Measurement 2026; 258: 119390. https://doi.org/10.1016/j.measurement.2025.119390.
  17. Li J, Kang X. Mobile-YOLO: An Accurate and Efficient Three-Stage Cascaded Network for Online Fiberglass Fabric Defect Detection. Engineering Applications of Artificial Intelligence 2024; 134: 108690. https://doi.org/10.1016/j.engappai.2024.108690.
  18. Wang Z, Zhou W, Li Y. MSAF-YOLO: An Efficient Multi-Scale Attention Fusion Network for High-Precision Steel Surface Defect Detection. Measurement 2026; 257: 118640. https://doi.org/10.1016/j.measurement.2025.118640.
  19. Qingyi W, Lingyun G, Bo C. The Lightweight CS-YOLO Model Applied for Concrete Crack Segmentation. Expert Systems with Applications 2025; 277: 127298. https://doi.org/10.1016/j.eswa.2025.127298.
  20. Yang R, Zhang H. Tunnel-YOLO: An Improved You Only Look Once Algorithm for Real-Time Shield Tunnel Lining Leakage Detection. Engineering Applications of Artificial Intelligence 2025; 162: 112403. https://doi.org/10.1016/j.engappai.2025.112403.
  21. Wu P, Xu Y, Ma Y, et al. LYA-YOLO: A Lightweight and Accurate YOLO Model in Drone Aerial Image Scenes. Expert Systems with Applications 2026; 321: 132166. https://doi.org/10.1016/j.eswa.2026.132166.
  22. Iqra, Giri KJ. SO-YOLOv8: A Novel Deep Learning-Based Approach for Small Object Detection with YOLO beyond COCO. Expert Systems with Applications 2025; 280: 127447. https://doi.org/10.1016/j.eswa.2025.127447.
  23. Hu J, Shen L, Sun G. Squeeze-and-Excitation Networks. In Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA, 18–23 June 2018. https://doi.org/10.1109/cvpr.2018.00745.
  24. Woo S, Park J, Lee JY, et al. CBAM: Convolutional Block Attention Module. In Computer Vision—ECCV 2018, Proceedings the 15th European Conference, Munich, Germany, 8–14 September 2018; Springer International Publishing: Cham, Switzerland, 2018. https://doi.org/10.1007/978-3-030-01234-2_1.
  25. Wang Q, Wu B, Zhu P, et al. ECA-Net: Efficient Channel Attention for Deep Convolutional Neural Networks. In Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 13–19 June 2020. https://doi.org/10.1109/cvpr42600.2020.01155.
  26. Hou Q, Zhou D, Feng J. Coordinate Attention for Efficient Mobile Network Design. In Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA, 20–25 June 2021. https://doi.org/10.1109/cvpr46437.2021.01350.
  27. Yang L, Zhang RY, Li L, et al. SimAM: A Simple, Parameter-Free Attention Module for Convolutional Neural Networks. In Proceedings of the 38th International Conference on Machine Learning, Virtual, 18–24 July 2021; pp. 11863–11874.
  28. Wei C, Wang W, Yang W, et al. Deep Retinex Decomposition for Low-Light Enhancement. In Proceedings of the British Machine Vision Conference (BMVC 2018), Tyne, UK, 3–6 September 2018.
  29. Guo C, Li C, Guo J, et al. Zero-Reference Deep Curve Estimation for Low-Light Image Enhancement. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 13–19 June 2020. https://doi.org/10.1109/cvpr42600.2020.00185.
  30. Jiang Y, Gong X, Liu D, et al. EnlightenGAN: Deep Light Enhancement without Paired Supervision. IEEE Transactions on Image Processing 2021; 30: 2340–2349. https://doi.org/10.1109/tip.2021.3051462.
  31. Liu R, Ma L, Zhang J, et al. Retinex-Inspired Unrolling with Cooperative Prior Architecture Search for Low-Light Image Enhancement. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA, 20–25 June 2021. https://doi.org/10.1109/cvpr46437.2021.01042.
  32. Wang Y, Wan R, Yang W, et al. Low-Light Image Enhancement with Normalizing Flow. Proceedings of the AAAI Conference on Artificial Intelligence 2022; 36: 2604–2612. https://doi.org/10.1609/aaai.v36i3.20162.
  33. Ren Z, Lin T, Feng K, et al. A Systematic Review on Imbalanced Learning Methods in Intelligent Fault Diagnosis. IEEE Transactions on Instrumentation and Measurement 2023; 72: 1–35. https://doi.org/10.1109/tim.2023.3246470.
  34. Fu T, Li P, Liu S. An Imbalanced Small Sample Slab Defect Recognition Method Based on Image Generation. Journal of Manufacturing Processes 2024; 118: 376–388. https://doi.org/10.1016/j.jmapro.2024.03.028.
  35. Lin TY, Goyal P, Girshick R, et al. Focal Loss for Dense Object Detection. In Proceedings of the 2017 IEEE International Conference on Computer Vision (ICCV), Venice, Italy, 22–29 October 2017. https://doi.org/10.1109/iccv.2017.324.
  36. Cui Y, Jia M, Lin TY, et al. Class-Balanced Loss Based on Effective Number of Samples. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, CA, USA, 15–20 June 2019. https://doi.org/10.1109/cvpr.2019.00949.
  37. Cao K, Wei C, Gaidon A, et al. Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss. In Proceedings of the Advances in Neural Information Processing Systems (NeurIPS 2019), Vancouver, BC, Canada, 8–14 December 2019.
  38. Kang B, Xie S, Rohrbach M, et al. Decoupling Representation and Classifier for Long-Tailed Recognition. In Proceedings of the International Conference on Learning Representations (ICLR 2020), Virtual, 26 April–1 May 2020.
  39. Ren J, Yu C, Sheng S, et al. Balanced Meta-Softmax for Long-Tailed Visual Recognition. In Proceedings of the Advances in Neural Information Processing Systems (NeurIPS 2020), Virtual, 6–12 December 2020; pp. 4175–4186.
  40. Menon AK, Jayasinghe S, Rawat AS, et al. Long-Tail Learning via Logit Adjustment. In Proceedings of the International Conference on Learning Representations (ICLR 2021), Virtual, 3–7 May 2021.
  41. Tan J, Wang C, Li B, et al. Equalization Loss for Long-Tailed Object Recognition. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 13–19 June 2020. https://doi.org/10.1109/cvpr42600.2020.01168.
  42. Wang J, Zhang W, Zang Y, et al. Seesaw Loss for Long-Tailed Instance Segmentation. In Proceedings of the 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA, 20–25 June 2021. https://doi.org/10.1109/cvpr46437.2021.00957.
  43. Sandler M, Howard A, Zhu M, et al. MobileNetV2: Inverted Residuals and Linear Bottlenecks. In Proceedings of the 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, USA, 18–23 June 2018. https://doi.org/10.1109/cvpr.2018.00474.
  44. Ma N, Zhang X, Zheng HT, et al. ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design. In Computer Vision—ECCV 2018, Proceedings of the 15th European Conference, Munich, Germany, 8–14 September 2018; Springer International Publishing: Cham, Switzerland, 2018. https://doi.org/10.1007/978-3-030-01264-9_8.
  45. Han K, Wang Y, Tian Q, et al. GhostNet: More Features from Cheap Operations. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 13–19 June 2020. https://doi.org/10.1109/cvpr42600.2020.00165.
  46. Zhao Y, Lv W, Xu S, et al. DETRs Beat YOLOs on Real-Time Object Detection. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, 16–22 June 2024. https://doi.org/10.1109/cvpr52733.2024.01605.

Supporting Agencies

  1. Funding: This research was funded by Yongchuan District Science and Technology Projects, grant number (2024yc-jbgs20012).